CPRI Specification V4.

2 (2010-09-29)
Interface Specification
Common Public Radio Interface (CPRI);
Interface Specification

The CPRI specification has been developed by Ericsson AB, Huawei Technologies Co. Ltd, NEC Corporation, Alcatel Lucent and Nokia Siemens
Networks GmbH & Co. KG (the Parties) and may be updated from time to time. Further information about CPRI, and the latest specification,
may be found at http://www.cpri.info

BY USING THE CPRI SPECIFICATION, YOU ACCEPT THE Interface Specification Download Terms and Conditions FOUND AT
http://www.cpri.info/spec.html

IN ORDER TO AVOID ANY DOUBT, BY DOWNLOADING AND/OR USING THE CPRI SPECIFICATION NO EXPRESS OR IMPLIED LICENSE
AND/OR ANY OTHER RIGHTS WHATSOEVER ARE GRANTED FROM ANYBODY.

2009 Ericsson AB, Huawei Technologies Co. Ltd, NEC Corporation, Alcatel Lucent, and Nokia Siemens Networks GmbH & Co. KG.
.

CPRI
CPRI Specification V4.2 (2010-09-29) 2
Table of Contents
1. Introduction ...................................................................................................................4
2. System Description.......................................................................................................6
2.1. Definitions/Nomenclature...............................................................................6
2.2. System Architecture........................................................................................9
2.3. Reference Configurations.............................................................................10
2.4. Functional Description..................................................................................13
2.4.1. Radio Functionality ............................................................................13
2.4.2. CPRI Control Functionality ................................................................14
3. Interface Baseline........................................................................................................15
3.1. Supported Radio Standards .........................................................................15
3.2. Operating Range............................................................................................15
3.3. Topology/Switching/Multiplexing ................................................................15
3.4. Bandwidth/Capacity/Scalability ...................................................................17
3.4.1. Capacity in terms of Antenna-Carriers...............................................17
3.4.2. Required U-plane IQ Sample Widths.................................................18
3.4.3. Required C&M-plane Bit Rate ...........................................................19
3.5. Synchronization/Timing................................................................................19
3.5.1. Frequency Synchronization...............................................................19
3.5.2. Frame Timing Information .................................................................20
3.5.3. Link Timing Accuracy ........................................................................21
3.5.4. Round Trip Delay Accuracy...............................................................21
3.5.5. Accuracy of TDD Tx-Rx switching point ............................................22
3.6. Delay Calibration ...........................................................................................22
3.6.1. Round Trip Cable Delay per Link ......................................................22
3.6.2. Round Trip Delay of a Multi-hop Connection.....................................23
3.7. Link Maintenance ..........................................................................................23
3.8. Quality of Service ..........................................................................................24
3.8.1. Maximum Delay.................................................................................24
3.8.2. Bit Error Ratio U-plane ......................................................................24
3.8.3. Bit Error Ratio C&M-plane.................................................................24
3.9. Start-up Requirement ....................................................................................25
3.9.1. Clock Start-up Time Requirement .....................................................25
3.9.2. Plug and Play Requirement ...............................................................25
4. Interface Specification ................................................................................................27
4.1. Protocol Overview.........................................................................................27
4.2. Physical Layer (Layer 1) Specification ........................................................28
4.2.1. Line Bit Rate......................................................................................28
4.2.2. Physical Layer Modes .......................................................................28
4.2.3. Electrical Interface.............................................................................30
4.2.4. Optical Interface ................................................................................30
4.2.5. Line Coding .......................................................................................30
4.2.6. Bit Error Correction/Detection............................................................30
4.2.7. Frame Structure.................................................................................30
4.2.8. Synchronisation and Timing ..............................................................55
4.2.9. Link Delay Accuracy and Cable Delay Calibration ............................56
4.2.10. Link Maintenance of Physical Layer ..................................................59
4.3. Data Link Layer (Layer 2) Specification for Slow C&M Channel ...............64
4.3.1. Layer 2 Framing ................................................................................64
4.3.2. Media Access Control/Data Mapping ................................................64
4.3.3. Flow Control ......................................................................................65
4.3.4. Control Data Protection/ Retransmission Mechanism.......................65
4.4. Data Link Layer (Layer 2) Specification for Fast C&M Channel ................65

CPRI
CPRI Specification V4.2 (2010-09-29) 3
4.4.1. Layer 2 Framing ................................................................................65
4.4.2. Media Access Control/Data Mapping ................................................66
4.4.3. Flow Control ......................................................................................68
4.4.4. Control Data Protection/ Retransmission Mechanism.......................68
4.5. Start-up Sequence.........................................................................................68
4.5.1. General ..............................................................................................68
4.5.2. Layer 1 Start-up Timer.......................................................................69
4.5.3. State Description ...............................................................................70
4.5.4. Transition Description........................................................................74
5. Interoperability ............................................................................................................77
5.1. Forward and Backward Compatibility .........................................................77
5.1.1. Fixing Minimum Control Information Position in CPRI Frame
Structure............................................................................................77
5.1.2. Reserved Bandwidth within CPRI......................................................77
5.1.3. Version Number.................................................................................77
5.1.4. Specification Release Version mapping into CPRI Frame ................77
5.2. Compliance ....................................................................................................78
6. Annex ...........................................................................................................................79
6.1. Delay Calibration Example (Informative).....................................................79
6.2. Electrical Physical Layer Specification (Informative) ................................82
6.2.1. Overlapping Rate and Technologies .................................................82
6.2.2. Signal Definition.................................................................................83
6.2.3. Eye Diagram and Jitter ......................................................................83
6.2.4. Reference Test Points .......................................................................84
6.2.5. Cable and Connector.........................................................................84
6.2.6. Impedance.........................................................................................84
6.2.7. AC Coupling ......................................................................................84
6.2.8. TX Performances...............................................................................85
6.2.9. Receiver Performances .....................................................................90
6.2.10. Measurement Procedure...................................................................93
6.3. Networking (Informative) ..............................................................................94
6.3.1. Concepts ...........................................................................................94
6.3.2. Reception and Transmission of SAP
CM
by the RE.............................94
6.3.3. Reception and Transmission of SAP
IQ
by the RE..............................95
6.3.4. Reception and Distribution of SAP
S
by the RE..................................95
6.3.5. Reception and Transmission of CPRI Layer 1 Signalling by the
RE......................................................................................................95
6.3.6. Bit Rate Conversion...........................................................................95
6.3.7. More than one REC in a radio base station.......................................95
6.3.8. The REC as a Networking Element ...................................................96
6.4. E-UTRA sampling rates (Informative)..........................................................96
6.5. Scrambling (Normative) ................................................................................96
6.5.1. Transmitter.........................................................................................97
6.5.2. Receiver ..........................................................................................100
7. List of Abbreviations.................................................................................................101
8. References .................................................................................................................104
9. History ........................................................................................................................105


CPRI
CPRI Specification V4.2 (2010-09-29) 4
1. Introduction
The Common Public Radio Interface (CPRI) is an industry cooperation aimed at defining a publicly available
specification for the key internal interface of radio base stations between the Radio Equipment Control (REC)
and the Radio Equipment (RE). The parties cooperating to define the specification are Ericsson AB, Huawei
Technologies Co. Ltd, NEC Corporation, Alcatel Lucent and Nokia Siemens Networks GmbH & Co. KG.

Motivation for CPRI:
The CPRI specification enables flexible and efficient product differentiation for radio base stations and
independent technology evolution for Radio Equipment (RE) and Radio Equipment Control (REC).

Scope of Specification:
The necessary items for transport, connectivity and control are included in the specification. This includes
User Plane data, Control and Management Plane transport mechanisms, and means for synchronization.
A focus has been put on hardware dependent layers (layer 1 and layer 2). This ensures independent
technology evolution (on both sides of the interface), with a limited need for hardware adaptation. In addition,
product differentiation in terms of functionality, management, and characteristics is not limited.
With a clear focus on layer 1 and layer 2 the scope of the CPRI specification is restricted to the link interface
only, which is basically a point to point interface. Such a link shall have all the features necessary to enable a
simple and robust usage of any given REC/RE network topology, including a direct interconnection of multi-
port REs.
Redundancy mechanisms are not described in the CPRI specification, however all the necessary features to
support redundancy, especially in system architectures providing redundant physical interconnections (e.g.
rings) are defined.

CPRI
CPRI Specification V4.2 (2010-09-29) 5
The specification has the following scope (with reference to Figure 1):
1. A digitized and serial internal radio base station interface that establishes a connection between
Radio Equipment Control (REC) and Radio Equipment (RE) enabling single-hop and multi-hop
topologies is specified.
1

2. Three different information flows (User Plane data, Control and Management Plane data, and
Synchronization Plane data) are multiplexed over the interface.
3. The specification covers layers 1 and 2.
3a. The physical layer (layer 1) supports both an electrical interface (e.g., what is used in traditional
radio base stations), and an optical interface (e.g. for radio base stations with remote radio
equipment).
3b. Layer 2 supports flexibility and scalability.



Figure 1: System and Interface Definition

1
The CPRI specification may be used for any internal radio base station interface that carries the information flows mentioned in the
scope of point 2.
Radio Equipment (RE) Radio Equipment Control (REC)
Layer 1
Layer 2
Control &
Mgmt.
User Sync.
Air
Interface
Network
Interface
DigitizedRadio Base Station
Internal Interface Specification
Layer 1
Layer 2
Control &
Mgmt.
User Sync.

CPRI
CPRI Specification V4.2 (2010-09-29) 6
2. System Description
This chapter describes the CPRI related parts of the basic radio base station system architecture and defines
the mapping of the functions onto the different subsystems. Furthermore, the reference configurations and
the basic nomenclature used in the following chapters are defined.
The following description is based on the UMTS (Universal Mobile Telecommunication System), WiMAX
Forum Mobile System Profile [11] based on IEEE Std 802.16-2009 [13] and Evolved UMTS Terrestrial Radio
Access (E-UTRA). However, the interface may also be used for other radio standards.
2.1. Definitions/Nomenclature
This section provides the basic nomenclature that is used in the following chapters.
Subsystems:
The radio base station system is composed of two basic subsystems, the radio equipment control and the
radio equipment (see Figure 1). The radio equipment control and the radio equipment are described in the
following chapter.
Node:
The subsystems REC and RE are also called nodes, when either an REC or an RE is meant. The Radio
Base Station system shall contain at least two nodes, at least one of each type; REC and RE.
Protocol layers:
This specification defines the protocols for the physical layer (layer 1) and the data link layer (layer 2).
Layer 1 defines:
Electrical characteristics
Optical characteristics
Time division multiplexing of the different data flows
Low level signalling
Layer 2 defines:
Media access control
Flow control
Data protection of the control and management information flow
Protocol data planes:
The following data flows are discerned:
Control Plane: Control data flow used for call processing.
Management Plane: This data is management information for the operation, administration and
maintenance of the CPRI link and the nodes.
User Plane: Data that has to be transferred from the radio base station to the mobile station and
vice versa. These data are transferred in the form of IQ data.
Synchronization: Data flow which transfers synchronization and timing information between nodes.
The control plane and management plane are mapped to a Service Access Point SAP
CM
as described below.
User plane data:

The user plane data is transported in the form of IQ data. Several IQ data flows are sent via one physical
CPRI link. Each IQ data flow reflects the data of one antenna for one carrier, the so-called antenna-carrier
(AxC).


CPRI
CPRI Specification V4.2 (2010-09-29) 7
Antenna-carrier (AxC):
One antenna-carrier is the amount of digital baseband (IQ) U-plane data necessary for either reception or
transmission of only one carrier at one independent antenna element.
Antenna-carrier (AxC) Group:
An AxC Group is an aggregation of N
A
AxC with the same sample rate, the same sample width, the same
destination SAP
IQ
, and the same radio frame length. In case of N
A
=1 an AxC Group is the same as an AxC.
AxC Container:
An AxC Container is a sub-part of the IQ-data block of one basic frame. For UTRA-FDD it contains the IQ
samples of one AxC for the duration of one UMTS chip. For WiMAX it contains IQ sample bits of one AxC
and sometimes also stuffing bits. For E-UTRA it contains one or more IQ samples for the duration of one
UMTS chip or it contains IQ sample bits and sometimes also stuffing bits.
AxC Container Group:
An AxC Container Group is an aggregation of N
C
AxC Containers containing IQ-samples for an AxC
Group in one basic frame. N
C
is defined in section 4.2.7.2.7.
AxC Symbol Block:
An AxC Symbol Block is an aggregation in time of N
SAM
IQ samples for one WiMAX symbol plus N
S_SYM

stuffing bits. N
SAM
and N
S_SYM
are defined in section 4.2.7.2.6.
AxC Container Block:
An AxC Container Block is an aggregation in time of K AxC Container Groups or an aggregation in time of
N
SYM
AxC Symbol Blocks plus N
S_FRM
stuffing bits. It contains S IQ samples per AxC plus stuffing bits. K
and S are defined in section 4.2.7.2. N
SYM
and N
S_FRM
are defined in section 4.2.7.2.6.
Service Access Points:
For all protocol data planes, layer 2 service access points are defined that are used as reference points for
performance measurements. These service access points are denoted as SAP
CM
, SAP
S
and SAP
IQ
as
illustrated in Figure 2. A service access point is defined on a per link basis.
Stuffing bits:
Stuffing bits are used for alignment of WiMAX/E-UTRA sample frequencies to the basic frame frequency.
Stuffing bits are also sent in TDD mode during time intervals when there is no IQ data to be sent over CPRI.
The content of stuffing bits is vendor specific (v).
Stuffing samples:
If the total sampling rate per AxC Group is not the integer multiple of the CPRI basic frame rate (3.84MHz),
then stuffing samples are added to make the total sampling rate the integer multiple of the CPRI basic frame
rate. Stuffing samples are filled with vendor specific bits (v).
Link:
The term link is used to indicate the bidirectional interface in between two directly connected ports, either
between REC and RE, or between two nodes, using one transmission line per direction. A working link
consists of a master port, a bidirectional cable, and a slave port.
Master/master and slave/slave links are not covered by this specification (for the definition of master and
slave see below).
Passive Link:
A passive link does not support any C&M channel, i.e. it carries only IQ data and synchronization
information. It may be used for capacity expansion or redundancy purposes, or for any other internal
interfaces in a radio base station.
Hop:
A hop is the aggregation of all links directly connecting two nodes.
Multi-hop connection:
A multi-hop connection is composed of a set of continuously connected hops starting from the REC and
ending at a particular RE including nodes in between.

CPRI
CPRI Specification V4.2 (2010-09-29) 8
Logical connection:
A logical connection defines the interconnection between a particular SAP (e.g., SAP
CM
) belonging to a port
of the REC and the corresponding peer SAP (e.g., SAP
CM
) belonging to a port of one particular RE and
builds upon a single hop, or a multi-hop connection, between the REC and that particular RE. Logical
connections for C&M data, user plane data and synchronization can be distinguished.
Master port and slave port:
Each link connects two ports which have asymmetrical functions and roles: a master and a slave.
This is implicitly defined in CPRI release 1 with the master port in the REC and the slave port in the RE.
This master/slave role split is true for the following set of flows of the interface:
Synchronization
C&M channel negotiation during start-up sequence
Reset indication
Start-up sequence
Such a definition allows the reuse of the main characteristic of the CPRI release 1 specification, where each
link is defined with one termination being the master port and the other termination being the slave port.
At least one REC in a radio base station shall have at least one master port and optionally have other ports
that may be slave or master.
An RE shall have at least one slave port and optionally have other ports that may be slave or master.
Under normal conditions a link has always one master port and one slave port. Two master ports or two
slave ports connected together is an abnormal situation and is therefore not covered by this specification.
Downlink:
Direction from REC to RE for a logical connection.
Uplink:
Direction from RE to REC for a logical connection.

Figure 1A and Figure 1B illustrate some of the definitions.

REC RE #1 RE #2
Master
Port
Slave
Port
Link
Logical Connection for IQ data (RECRE #2)
Logical Connection for Synchronization (RECRE #2)
Logical Connection for C&M data (RECRE #2)
Master
Port
Slave
Port
SAP
CM
SAP
IQ
Hop
SAP
S
SAP
CM
SAP
CM
SAP
IQ
SAP
IQ
SAP
S
SAP
S

Figure 1A: Illustration of basic definitions

CPRI
CPRI Specification V4.2 (2010-09-29) 9

Figure 1B: Illustration of AxC related definitions
2.2. System Architecture
Radio base stations should provide deployment flexibility for the mobile network operators, i.e., in addition to
a concentrated radio base station, more flexible radio base station system architectures involving remote
radio equipment shall be supported. This may be achieved by a decomposition of the radio base station into
two basic building blocks, the so-called radio equipment control (REC) and the radio equipment (RE) itself.
Both parts may be physically separated (i.e., the RE may be close to the antenna, whereas the REC is
located in a conveniently accessible site) or both may be co-located as in a conventional radio base station
design.
The REC contains the radio functions of the digital baseband domain, whereas the RE contains the
analogue radio frequency functions. The functional split between both parts is done in such a way that a
generic interface based on In-Phase and Quadrature (IQ) data can be defined.
For the UMTS radio access network, the REC provides access to the Radio Network Controller via the Iub
interface, whereas the RE serves as the air interface, called the Uu interface, to the user equipment.
For WiMAX, the REC provides access to network entities (e.g. other BS, ASN-GW), whereas the RE serves
as the air interface to the subscriber station / mobile subscriber station (SS / MSS).
For E-UTRA, the REC provides access to the Evolved Packet Core for the transport of user plane and
control plane traffic via S1 interface, whereas the RE serves as the air interface to the user equipment.
A more detailed description of the functional split between both parts of a radio base station system is
provided in Section 2.4.
In addition to the user plane data (IQ data), control and management as well as synchronization signals have
to be exchanged between the REC and the RE. All information flows are multiplexed onto a digital serial
communication line using appropriate layer 1 and layer 2 protocols. The different information flows have
access to the layer 2 via appropriate service access points. This defines the common public radio interface
illustrated in Figure 2. The common public radio interface may also be used as a link between two nodes in
system architectures supporting networking. An example of a common public radio interface between two
REs is illustrated in Figure 2A.
AxC
AxC
Basic Frame
AxC
Container
AxC
container
AxC
AxC
Container
K Basic Frames, S samples
AxC
Container
AxC
Container
AxC
Container
AxC Symbol Block
AxC
Container
Group
WiMAX Symbol
AxC
Group
SAP
IQ
SAP
IQ
AxC Container Block
WiMAX Symbol
WiMAX Frame
time
AxC
AxC
Basic Frame
AxC
Container
AxC
Container
AxC
container
AxC
AxC
Container
K Basic Frames, S samples
AxC Container Block
AxC
Container
AxC
Container
AxC
Container
AxC
Container
AxC
Container
AxC
Container
AxC
Container
Group
WiMAX Symbol
AxC
SAP
IQ
SAP
IQ
AxC Symbol Block
AxC Container Block
WiMAX Symbol
WiMAX Frame

CPRI
CPRI Specification V4.2 (2010-09-29) 10

Figure 2: Basic System Architecture and Common Public Radio Interface Definition

Figure 2A: System Architecture with a link between REs
2.3. Reference Configurations
This section provides the reference configurations that have to be supported by the CPRI specification. The
basic configuration, shown in Figure 3, is composed of one REC and one RE connected by a single CPRI
link. The basic configuration can be extended in several ways:
First, several CPRI links may be used to enhance the system capacity as required for large system
configurations involving many antennas and carriers (see Figure 4). It is required that an IQ data flow
of a certain antenna and a certain antenna-carrier (see Section 2.1) is carried completely by one
CPRI link (however, it is allowed that the same antenna-carrier may be transmitted simultaneously
over several links). Therefore, the number of physical links is not restricted by this specification.
Second, several REs may be served by one REC as illustrated in Figure 5 for the so-called star
topology.
Third, one RE may be served by multiple RECs as illustrated in Figure 5D. The requirements for this
configuration are not fully covered in the CPRI specification; refer to section 6.3.7 for further
explanation.
Furthermore, three basic networking topologies may be used for the interconnection of REs:
o Chain topology, an example is shown in Figure 5A
o Tree topology, an example is shown in Figure 5B
o Ring topology, an example is shown in Figure 5C
Any other topology (e.g. combination of RECs and REs in a chain and tree) is not precluded. An
example of reusing the CPRI interface for other internal interfaces in a radio base station is depicted
in Figure 5E.

Radio Equipment (RE) #1 Radio Equipment Control (REC)
Layer 1
Layer 2
Control &
Mgmt
Sync
SAP
CM
Network
Interface
Common Public Radio Interface
Layer 1
Layer 2
Radio Base Station System
SAP
S
SAP
IQ
SAP
CM
SAP
S
SAP
IQ
Radio Equipment (RE) #2
Ai r Interface
Common Public Radio Interface
Layer 1
Layer 2
SAP
CM
SAP
S
SAP
IQ
Ai r Interface
Layer 1
Layer 2
SAP
CM
SAP
S
SAP
IQ
Control & Control & Control &
Mgmt Mgmt Mgmt
Sync Sync Sync User Plane User Plane User Plane User Plane
Radio Equipment (RE) Radio Equipment Control (REC)
Layer 1
Layer 2
Control &
Mgmt
User Plane Sync
SAP
CM
Air Interface
Network Interface
Common Public Radio Interface
Layer 1
Layer 2
Control &
Mgmt
User Plane Sync
Radio Base Station System
CPRI link
SAP
S
SAP
IQ
SAP
CM
SAP
S
SAP
IQ
Master port Slave port
Master port Master port Slave port Slave port
CPRI link CPRI link

CPRI
CPRI Specification V4.2 (2010-09-29) 11
o If a radio base station has multiple RECs, e.g. of different radio access technologies, the
CPRI interface may be used for the interface between two RECs.
o The requirements for this configuration are not fully covered in the CPRI specification; refer
to sections 6.3.7 and 6.3.8 for further explanation.

REC RE CPRI link

Figure 3: Single point-to-point link between one REC and one RE


REC RE
.
.
.
CPRI link
CPRI link

Figure 4: Multiple point-to-point links between one REC and one RE


REC
RE
RE
.
.
.
C
P
R
I

l
i
n
k
(
s
)
.
.
.
C
P
R
I

l
i
n
k
(
s
)

Figure 5: Multiple point-to-point links between one REC and several REs (star topology)



Figure 5A: Chain topology

REC RE CPRI link(s)
RE CPRI link(s)
CPRI link
...

CPRI
CPRI Specification V4.2 (2010-09-29) 12

Figure 5B: Tree topology


Figure 5C: Ring topology


Figure 5D: Multiple point-to-point links between several RECs and one RE


Figure 5E: Chain topology of multiple RECs

REC REC CPRI link(s)
RE CPRI link(s)
...
REC
CPRI link(s)
RE
CPRI link(s)
REC
RE
RE
...
...
CPRI link(s)
REC
RE
CPRI link(s)
CPRI link(s)
REC RE CPRI link(s)
RE CPRI link(s)
CPRI link(s)

CPRI
CPRI Specification V4.2 (2010-09-29) 13
2.4. Functional Description
2.4.1. Radio Functionality
This section provides a more detailed view on the functional split between REC and RE, which provides the
basis for the requirement definition in the next chapter.
The REC is concerned with the Network Interface transport, the radio base station control and management
as well as the digital baseband processing. The RE provides the analogue and radio frequency functions
such as filtering, modulation, frequency conversion and amplification. An overview on the functional
separation between REC and RE is given in Table 1 for UTRA FDD and in Table 1A for WiMAX and E-
UTRA.
Table 1: Functional decomposition between REC and RE (valid for the UTRA FDD standard)
Functions of REC Functions of RE
Downlink Uplink Downlink Uplink
Radio base station control & management
Iub transport RRC Channel Filtering
Iub Frame protocols D/A conversion A/D conversion
Channel Coding Channel De-coding Up Conversion Down Conversion
Interleaving De-Interleaving ON/OFF control of each
carrier
Automatic Gain Control
Spreading De-spreading Carrier Multiplexing Carrier De-multiplexing
Scrambling De-scrambling
MIMO processing
Power amplification and
limiting
Low Noise Amplification
Adding of physical
channels
Signal distribution to
signal processing units
Antenna supervision
Transmit Power Control
of each physical channel
Transmit Power Control &
Feedback Information
detection
RF filtering RF filtering
Frame and slot signal
generation (including
clock stabilization)

Measurements Measurements


CPRI
CPRI Specification V4.2 (2010-09-29) 14
Table 1A: Functional decomposition between REC and RE (valid for WiMAX & E-UTRA)
Functions of REC Functions of RE
Downlink Uplink Downlink Uplink
Radio base station control & management Add CP (optional)
Backhaul transport Channel Filtering
MAC layer D/A conversion A/D conversion
Channel Coding,
Interleaving, Modulation
Channel De-coding, De-
Interleaving,
Demodulation
Up Conversion Down Conversion
iFFT FFT ON/OFF control of each
carrier
Automatic Gain Control
Add CP (optional) Remove CP Carrier Multiplexing Carrier De-multiplexing
MIMO processing
Power amplification and
limiting
Low Noise Amplification
Signal aggregation from
signal processing units
Signal distribution to
signal processing units
Antenna supervision
Transmit Power Control
of each physical channel
Transmit Power Control &
Feedback Information
detection
RF filtering RF filtering
Frame and slot signal
generation (including
clock stabilization)
TDD switching
in case of TDD mode
Measurements Measurements

2.4.2. CPRI Control Functionality
This section provides a more detailed view on the functional split between REC and RE for CPRI
functionality beyond the specification itself.
Basically, the REC is concerned with the management of the CPRI and the CPRI topology. The RE may
optionally provide interconnection functionality between REs. An overview of the functional separation
between REC and RE is given in Table 1B.
Table 1B: Functional decomposition between REC and RE (valid for CPRI control functionality)
Functions of REC Functions of RE
Downlink Uplink Downlink Uplink
CPRI control management
CPRI topology management CPRI interconnection between REs
(forwarding/ switching/cross-connecting of CPRI SAP
data between REs)



CPRI
CPRI Specification V4.2 (2010-09-29) 15
3. Interface Baseline
This chapter provides input requirements for the CPRI specification. The requirements are to be met by the
CPRI specification, and will be used as a baseline for future enhancements of the CPRI specification. Note
that this chapter does not specify the requirements on a CPRI compliant device (see chapter 5.2) but
expresses the superset of requirements for an interface from all expected applications using the CPRI.
3.1. Supported Radio Standards
The interface shall support transmission of all necessary data between REC and RE in both directions for a
radio base station consisting of one REC and one or more REs compliant to the following radio standards:

Requirement No. Requirement Definition Requirement Value Scope
3GPP UTRA FDD, Release 9,
March 2010
WiMAX Forum Mobile System
Profile Release 1.5 Approved
Specification (2009-08-01)
R-1 Supported Radio
Standards and Releases
3GPP E-UTRA, Release 9,
March 2010
Logical connection

The support of other standards is not required in this release of the CPRI specification, but the future use of
the interface for other standards shall not be precluded.
3.2. Operating Range
The interface shall support a continuous range of distances (i.e., cable lengths) between master and slave
ports. The minimum required range is defined by the cable length in the following table:

Requirement No. Requirement Definition Requirement Value Scope
R-2 Cable length (lower limit) 0 m Link
R-3 Cable length (upper limit) >10 km Link

The interface shall support one cable between master and slave with separate transmission media (e.g.,
optical fibres) for uplink and downlink.
3.3. Topology/Switching/Multiplexing
The interface shall support the following networking topologies:

Requirement No. Requirement Definition Requirement Value Scope
R-4 Topology Star topology,
Chain topology,
Tree topology,
Ring topology
Radio
base
station
system


CPRI
CPRI Specification V4.2 (2010-09-29) 16

The support of other topologies is not required in this release of the specification, but the use of the interface
in other topologies shall not be precluded.

The interface shall support multiple hops when used in a networking configuration:

Requirement No. Requirement Definition Requirement Value Scope
R-4A Maximum number of hops
in a logical connection
At least 5 hops Logical connection

One RE may support several ports to fit in the different topologies but at least one is a slave port:

Requirement No. Requirement Definition Requirement Value Scope
R-4B Number of ports per RE RE may support
more than one CPRI
port
Node

Requirement No. Requirement Definition Requirement Value Scope
R-4C Number of slave ports
per RE
RE shall support at
least one CPRI
slave port
Node

A logical connection may use a multi-hop connection composed of links with different line bit rates.

Requirement No. Requirement Definition Requirement Value Scope
R-4D One logical connection
may consist of
successive hops with
different link numbers
and line bit rates.

N/A Logical
connection


It shall be possible to use a link as a redundant link in any network topology.

Requirement No. Requirement Definition Requirement Value Scope
R-4E A link may be used as a
redundant link in any
network topology.

N/A Link

It shall be possible to mix different Radio Standards on a link.

CPRI
CPRI Specification V4.2 (2010-09-29) 17

Requirement No. Requirement Definition Requirement Value Scope
R-4F Different Radio
Standards may be mixed
on a link.

N/A Link

3.4. Bandwidth/Capacity/Scalability
3.4.1. Capacity in terms of Antenna-Carriers
The capacity of one logical connection shall be expressed in terms of UTRA-FDD-antenna-carriers
(abbreviation: antenna-carrier or AxC). One UTRA-FDD-antenna-carrier is the amount of digital baseband
(IQ) U-plane data necessary for either reception or transmission of one UTRA-FDD carrier at one
independent antenna element. One antenna element is typically characterized by having exactly one
antenna connector to the RE.
CPRI shall be defined in such a way that the following typical Node B configurations can be supported:
1 RE supports one sector
o Up to 4 carriers x 1 antenna per RE (e.g. 6 REs for 3 sectors).
o Up to 4 carriers x 2 antennas per RE (e.g. 3 REs for 3 sectors)
1 RE supports 3 sectors
o From 1 to 4 carriers x 2 antennas x 3 sectors per RE


CPRI
CPRI Specification V4.2 (2010-09-29) 18
Therefore, the following number of AxC shall be supported by the CPRI specification:

Requirement No. Requirement Definition Requirement
Value
Scope
R-5 Number of antenna
carriers per logical
connection for UTRA FDD
only
4 Logical connection
R-6 Number of antenna
carriers per logical
connection for UTRA FDD
only
6 Logical connection
R-7 Number of antenna
carriers per logical
connection for UTRA FDD
only
8 Logical connection
R-8 Number of antenna
carriers per logical
connection for UTRA FDD
only
12 Logical connection
R-9 Number of antenna
carriers per logical
connection for UTRA FDD
only
18 Logical connection
R-10 Number of antenna
carriers per logical
connection for UTRA FDD
only
24 Logical connection

3.4.2. Required U-plane IQ Sample Widths
The IQ sample widths supported by the CPRI specification shall be between 4 and 20 bits for I and Q in the
uplink and between 8 and 20 bits in the downlink.

Requirement No. Requirement Definition Requirement
Value
Scope
R-11 Minimum uplink IQ sample
width for UTRA FDD only
4 Logical connection
R-11A Minimum uplink IQ sample
width for WiMAX and E-
UTRA
8 Logical connection
R-12 Maximum uplink IQ sample
width for UTRA FDD only
10 Logical connection
R-12A Maximum uplink IQ sample
width for WiMAX and E-
UTRA
20 Logical connection
R-13 Minimum downlink IQ
sample width
8 Logical connection
R-14 Maximum downlink IQ
sample width
20 Logical connection

CPRI
CPRI Specification V4.2 (2010-09-29) 19

Notes:
Oversampling Factor of 2 or 4 is assumed for UTRA FDD in uplink
Oversampling Factor of 1 or 2 is assumed for UTRA FDD in downlink.
Oversampling Factor of 1 is assumed for WiMAX and E-UTRA
Automatic Gain Control may be used in uplink

3.4.3. Required C&M-plane Bit Rate
The interface shall support a minimum bit rate for the M-plane transmission per link:
Requirement No. Requirement Definition Requirement Value Scope
R-15 Minimum transmission rate
of M-plane data (layer 1)
200 kbit/s Link

Additionally, the interface shall support a minimum bit rate for the transmission of C-plane data per AxC:
Requirement No. Requirement Definition Requirement Value Scope
R-16 Minimum transmission rate
of C-plane data (layer 1)
25 kbit/s Logical connection

The overhead on layer 2 due to frame delineation and frame check sequence depends on the frame length
determined by higher layers. Assuming this overhead is well below 20%, a minimum net bit rate of 20kbit/s
per AxC is available at the service access point SAP
CM
as shown in Figure 2 and Figure 2A.
3.5. Synchronization/Timing
3.5.1. Frequency Synchronization
The interface shall enable the RE to achieve the required frequency accuracy according to:
3GPP TS 25.104 [8] section 6.3 for UTRA FDD,
WiMAX Forum System Profile [11] section 4.2.4 for WiMAX
3GPP TS 36.104 [14], section 6.5.1 for E-UTRA.
The central clock for frequency generation in the RE shall be synchronized to the bit clock of one slave port.
With 8B/10B line coding the bit clock rate of the interface shall be a multiple of 38.4MHz in order to allow for
a simple synchronization mechanism and frequency generation in the RE.

The impact of jitter on the frequency accuracy budget of the interface to the radio base station depends on
the cut-off frequency of the RE synchronization mechanism. The interface shall accommodate a
synchronization mechanism cut-off frequency high enough so that a standard crystal oscillator suffices as
master clock of the RE. The contribution
0
f f of the jitter to the frequency accuracy shall be defined
with the cut-off frequency
CUT
f as follows:
df f
f f
f
dB
f L f
CUT
=

10
) (
0
2
0 0
10 2
1
, (1)
where ) ( f L is the single-side-band phase noise in dBc/Hz acquired on the interface with the following
relation to the jitter :

CPRI
CPRI Specification V4.2 (2010-09-29) 20
df
f
f
dB
f L


=

2
0
10
) (
0
0
10 2
2
1

(2)
The reference point for the jitter and phase noise specification is a stable clock signal at the service access
point SAP
S
as shown in Figure 2. The frequency of this clock signal is denoted as
0
f .
With
CUT
f in equation (1) being the maximum allowed cut-off frequency, the impact of jitter on the radio
base station frequency accuracy budget shall meet the following requirements:

Requirement No. Requirement Definition Requirement Value Scope
R-17 Maximum allowed cut-off
frequency
CUT
f of RE
synchronization
300 Hz Link
R-18 Maximum contribution
0
f f of jitter from the
CPRI link to the radio base
station frequency accuracy
budget (between master
SAP
S
and slave SAP
S
)
0.002 ppm Link


Any RE shall receive on its slave port a clock traceable to the main REC clock. This requires any RE reuses
on its master ports a transmit clock traceable to REC, i.e. a clock retrieved from one of its slave ports.

Requirement No. Requirement Definition Requirement Value Scope
R-18A Receive clock on RE slave
port
The clock shall be
traceable to REC clock
Link

Traceable clock means the clock is produced from a PLL chain system with REC clock as input. PLL
chain performance is out of CPRI scope.

3.5.2. Frame Timing Information
The synchronization part of the interface shall include mechanisms to provide precise frame timing
information from the REC to the RE. The frame timing information shall be recovered on the RE in order to
achieve the timing accuracy requirements as described in the sections below.
The RE shall forward frame timing information transparently when forwarding from a slave port to all the
master ports. The frame timing information is allocated to the service access point SAP
S
as shown in Figure
2. Timing accuracy and delay accuracy, as required in the subsections below, refer to the accuracy of timing
signals at the service access point SAP
S
. These timing signals shall be used in the RE for the precise timing
of RF signal transmission and reception on the air interface.


CPRI
CPRI Specification V4.2 (2010-09-29) 21
3.5.3. Link Timing Accuracy
In this section the link accuracy requirement (R-19) is introduced based on the following requirements from
the supported radio standards:
1. 3GPP UTRA-FDD Tx diversity and MIMO compliancy
2

The interface shall enable a radio base station to meet the requirement time alignment error in Tx
Diversity and MIMO transmission (3GPP TS 25.104 [8] section 6.8.4).
2. 3GPP UTRA-FDD UE positioning with GPS timing alignment:
The interface shall also support UTRAN GPS Timing of Cell Frames for UE positioning (3GPP TS
25.133 [9] section 9.2.10), which requires absolute delay accuracy.
3. WiMAX network synchronization with GPS (sections 8.3.7.1.1 and 8.4.10.1.1 of IEEE 802.16 [13])
4. E-UTRA Time alignment between transmitter branches
The interface shall enable a radio base station to meet the requirement time alignment between
transmitter branches (3GPP TS 36.104 [14], section 6.5.3).

Requirement R-19 is based on the following three criteria:
a) Meet the 1
st
and 4
th
requirement in a star configuration as shown in Figure 5 when TX diversity or MIMO
signals belonging to one cell are transmitted via different REs, where each RE is connected to the REC
via a single link.
b) Meet the 2
nd
and 3
rd
requirement at any RE connected to the REC via multi-hop connection to the REC
with the number of hops as given in R-4A.
c) Allow enough margin for additional delay tolerances in the RE implementation which is not part of CPRI.
The delay accuracy on one interface link excluding the group delay on the transmission medium, i.e.
excluding the cable length, shall meet the following requirement.

Requirement No. Requirement Definition Requirement Value Scope
R-19 Link delay accuracy in
downlink between SAP
S

master port and SAP
S

slave port excluding the
cable length.
8. 138ns
[ = T
C
/ 32]
Link

Note: The scope link for R-19 was chosen since the requirement R-19 can easily be met on a link, i.e. on a
single hop. In multi-hop configurations the delay tolerances per link may add up, so the total tolerance may
depend on the number of hops. Therefore it is not mandatory for CPRI to support a certain delay accuracy
requirement for all multi-hop connections.
3.5.4. Round Trip Delay Accuracy
The round trip delay accuracy requirement (R-20) is introduced based on the following requirements from the
supported radio standards:
3GPP UTRA-FDD, round trip time absolute accuracy
The interface shall enable a radio base station to meet the requirement round trip time absolute
accuracy 0.5 T
C
(3GPP TS 25.133 [9] section 9.2.8.1).
3GPP E-UTRA, timing advance
The interface shall enable a radio base station to meet the Timing Advance report mapping minimum
resolution of 65 ns (3GPP TS 36.133 [15], section 10.3).

2
With UTRA-FDD release 7, MIMO was introduced in the same section 6.8.4 of TS 25.104 [8] in addition to TX diversity without
changing the specification value.

CPRI
CPRI Specification V4.2 (2010-09-29) 22
The round trip time absolute accuracy of the interface, excluding the round trip group delay on the
transmission medium (i.e., excluding the cable length), shall meet the following requirement.

Requirement No. Requirement Definition Requirement Value Scope
R-20 Round trip absolute
accuracy excluding cable
length
16. 276ns
[ = T
C
/ 16]
Logical connection

Note: For round trip delay absolute accuracy even in multi-hop scenarios the delay tolerances per link do not
add up as can be seen from the timing relations in section 4.2.9 and annex 6.1. Therefore the scope of
requirement R-20 is logical connection, which can be met in all configurations.

3.5.5. Accuracy of TDD Tx-Rx switching point
For WiMAX and E-UTRA TDD applications the Tx Rx switching point needs to be transmitted per AxC. The
required maximum contribution of the interface to the switching point accuracy shall meet the following
requirement.

Requirement No. Requirement Definition Requirement Value Scope
R-20A Maximum contribution of
the interface to the
accuracy of TDD Tx-Rx
switching point
16. 276ns
[ = T
C
/ 16]
Multi-hop connection

3.6. Delay Calibration
3.6.1. Round Trip Cable Delay per Link
The interface shall enable periodic measurement of the cable length of each link, i.e., measurement of the
round trip group delay on the transmission medium of each link. The measurement results shall be available
on the REC in order to meet the following requirements without the need to input the cable length to the REC
by other means. The round trip delay accuracy requirement (R-21) is introduced based on the following
requirements from the supported radio standards:
time alignment error in Tx Diversity shall not exceed T
C
(3GPP TS 25.104 [8] section 6.8.4)
round trip time absolute accuracy 0.5 T
C
(3GPP TS 25.133 [9] section 9.2.8.1)
UTRAN GPS Timing of Cell Frames for UE positioning (3GPP TS 25.133 [9] section 9.2.10)
WiMAX network synchronization with GPS (sections 8.3.7.1.1 and 8.4.10.1.1 of IEEE 802.16 [13])
E-UTRA, Timing Advance minimum resolution of 65 ns (3GPP TS 36.133 [15], section 10.3)
The accuracy of the measurement of round trip group delay on the transmission medium of one link shall
meet the following requirement:

Requirement No. Requirement Definition Requirement Value Scope
R-21 Accuracy of the round trip
delay measurement of
cable delay of one link
16. 276ns
[ = T
C
/ 16]
Link


CPRI
CPRI Specification V4.2 (2010-09-29) 23
3.6.2. Round Trip Delay of a Multi-hop Connection
The interface shall enable periodic measurement of the round trip group delay of each multi-hop connection.
The measurement results shall be available on the REC in order to meet the following requirements without
the need to input the cable lengths of the involved links to the REC by other means. The round trip delay
accuracy requirement (R-21A) is introduced based on the following requirements from the supported radio
standards:
round trip time absolute accuracy 0.5 T
C
(3GPP TS 25.133 [9] section 9.2.8.1)
E-UTRA, Timing Advance minimum resolution of 65 ns (3GPP TS 36.133 [15] section 10.3)
By measuring the round trip delay of the multi-hop connection directly, REC based computation of round trip
delay shall be possible whatever the topology and the RE location within the branch, without adding delay
tolerances of all links and networking REs used in the multi-hop connection.
The accuracy of the measurement of round trip group delay on the multi-hop connection shall meet the
following requirement:

Requirement No. Requirement Definition Requirement Value Scope
R-21A Accuracy of the round trip
delay measurement of the
multi-hop connection
16. 276ns
[ = T
C
/ 16]
Multi-hop connection

3.7. Link Maintenance
The layer 1 of the interface shall be able to detect and indicate loss of signal (LOS) and loss of frame (LOF)
including frame synchronization. A remote alarm indication (RAI) shall be returned to the sender on layer 1
as a response to these errors. In addition the SAP defect indication (SDI) shall be sent to the remote end
when any of the service access points is not valid due to an equipment error.
The signals
LOS
LOF
SDI
RAI
shall be handled within layer 1 and shall also be available to the higher layers of the interface.

Requirement No. Requirement Definition Requirement Value Scope
R-22 Loss of Signal (LOS)
detection and indication
- Link
R-23 Loss of Frame (LOF)
detection and indication
- Link
R-24 SAP Defect Indication
(SDI)
- Link
R-25 Remote Alarm Indication
(RAI)
- Link


CPRI
CPRI Specification V4.2 (2010-09-29) 24
3.8. Quality of Service
3.8.1. Maximum Delay
In order to support efficient implementation of UTRA-FDD inner loop power control
3
, the absolute round trip
time for U-plane data (IQ data) on the interface, excluding the round trip group delay on the transmission
medium (i.e. excluding the cable length), shall not exceed the following maximum value:

Requirement No. Requirement Definition Requirement Value Scope
R-26 Maximum absolute round
trip delay per link
excluding cable length
5s Link

Round trip time is defined as the downlink delay plus the uplink delay. The delay is precisely defined as the
time required transmitting a complete IQ-sample over the interface. The availability and validity of an IQ-
sample is defined at the service access point SAP
IQ
as shown in Figure 2. The precise point of time of
availability and validity is indicated by the edge of an associated clock signal at the service access point
SAP
IQ
. The delay (e.g. in downlink) is defined as the time difference between the edge at the input SAP
IQ
(e.g. on REC or RE) and the edge at the output SAP
IQ
(e.g. on RE).
This definition is only valid for a regular transmission of IQ samples with a fixed sample clock.

3.8.2. Bit Error Ratio U-plane
The interface shall provide U-plane data transmission (on layer 1) with a maximum bit error ratio as specified
below:

Requirement No. Requirement Definition Requirement Value Scope
R-27 Maximum bit error ratio
(BER) of U-plane
10
-12

Link

It should be a design goal to avoid forward error correction on layer 1 to achieve a cost efficient solution.
There shall not be any data protection on layer 2.
3.8.3. Bit Error Ratio C&M-plane
The interface shall provide C&M-plane data transmission with a maximum bit error ratio (on layer 1) as
specified below:

Requirement No. Requirement Definition Requirement Value Scope
R-28 Maximum bit error ratio
(BER) of C&M-plane
10
-12

Link

Additionally, a frame check sequence (FCS) shall be provided for C&M-plane data bit error detection on layer
2. The minimum length of the frame check sequence is defined in the following table:

3
Even with the introduction of new standards (e.g. WiMAX and E-UTRA) UTRA FDD inner loop power control is still assumed to be the
most time critical procedure constraining R-26

CPRI
CPRI Specification V4.2 (2010-09-29) 25

Requirement No. Requirement Definition Requirement Value Scope
R-29 Minimum length of frame
check sequence (FCS)
16 bit Link

3.9. Start-up Requirement
3.9.1. Clock Start-up Time Requirement
CPRI shall enable the RE clock to achieve synchronization with respect to the frequency accuracy and
absolute frame timing accuracy within 10 seconds. The time needed for auto-negotiation of features (see
Plug and Play requirement in section 3.9.2) is excluded from this requirement.

Requirement No. Requirement Definition Requirement Value Scope
R-30 Maximum clock
synchronization time
10 s Link

3.9.2. Plug and Play Requirement
CPRI shall support auto-negotiation for selecting the line bit rate.

Requirement No. Requirement Definition Requirement Value Scope
R-31 Auto-negotiation of line bit rate - Link

CPRI shall support auto-negotiation for selecting the C&M-plane type and bit rate (layer 1).

Requirement No. Requirement Definition Requirement Value Scope
R-32 Auto-negotiation of C&M-plane
type and bit rate (layer 1)
- Link

CPRI shall support auto-detection of REC data flow on slave ports:

Requirement No. Requirement Definition Requirement Value Scope
R-33 Auto-detection of REC data flow
on slave ports
- Link

CPRI shall support auto-negotiation of scrambling:

Requirement No. Requirement Definition Requirement Value Scope
R-34 Auto-negotiation of scrambling - Link

CPRI shall support auto-detection of the scrambling seed:


CPRI
CPRI Specification V4.2 (2010-09-29) 26
Requirement No. Requirement Definition Requirement Value Scope
R-35 Auto-detection of scrambling
seed
- Link






CPRI
CPRI Specification V4.2 (2010-09-29) 27
4. Interface Specification
4.1. Protocol Overview
CPRI defines the layer 1 and layer 2 protocols for the transfer of user plane, C&M as well as synchronization
information between REC and RE as well as between two REs
4
. The interface supports the following types
of information flows:
IQ Data: User plane information in the form of in-phase and quadrature
modulation data (digital baseband signals).
Synchronization: Synchronization data used for frame and time alignment.
L1 Inband Protocol: Signalling information that is related to the link and is directly
transported by the physical layer. This information is required, e.g. for
system start-up, layer 1 link maintenance and the transfer of time
critical information that has a direct time relationship to layer 1 user
data.
C&M data: Control and management information exchanged between the control
and management entities within the REC and the RE. This information
flow is given to the higher protocol layers.
Protocol Extensions: This information flow is reserved for future protocol extensions. It may
be used to support, e.g., more complex interconnection topologies or
other radio standards.
Vendor Specific Information: This information flow is reserved for vendor specific information.
The user plane information is sent in the form of IQ data. The IQ data of different antenna carriers are
multiplexed by a time division multiplexing scheme onto an electrical or optical transmission line. The control
and management data are either sent as inband protocol (for time critical signalling data) or by layer 3
protocols (not defined by CPRI) that reside on top of appropriate layer 2 protocols. Two different layer 2
protocols for C&M data subset of High level Data Link Control (HDLC) and Ethernet are supported by
CPRI. These additional control and management data are time multiplexed with the IQ data. Finally,
additional time slots are available for the transfer of any type of vendor specific information. Figure 6
provides an overview on the basic protocol hierarchy.

4
The CPRI protocol may be reused for any internal radio base station interfaces.

CPRI
CPRI Specification V4.2 (2010-09-29) 28
Time Division Multiplexing
User Plane
Control&
Management
Plane
Electrical
Transmission
Optical
Transmission
IQ
Data
E
t
h
e
r
n
e
t
H
D
L
C
L
1

I
n
b
a
n
d

P
r
o
t
o
c
o
l
V
e
n
d
o
r

S
p
e
c
i
f
i
c
Layer 1
Layer 2
SYNC

Figure 6: CPRI protocol overview
4.2. Physical Layer (Layer 1) Specification
4.2.1. Line Bit Rate
In order to achieve the required flexibility and cost efficiency, several different line bit rates are defined.
Therefore, the CPRI line bit rate may be selected from the following option list:
CPRI line bit rate option 1: 614.4 Mbit/s
CPRI line bit rate option 2: 1228.8 Mbit/s (2 x 614.4 Mbit/s)
CPRI line bit rate option 3: 2457.6 Mbit/s (4 x 614.4 Mbit/s)
CPRI line bit rate option 4: 3072.0 Mbit/s (5 x 614.4 Mbit/s)
CPRI line bit rate option 5: 4915.2 Mbit/s (8 x 614.4 Mbit/s)
CPRI line bit rate option 6: 6144.0 Mbit/s (10 x 614.4 Mbit/s)
CPRI line bit rate option 7: 9830.4 Mbit/s (16 x 614.4 Mbit/s)
It is mandatory that each REC and RE support at least one of the above cited CPRI line bit rates.
All CPRI line bit rates have been chosen in such a way that the basic UMTS chip rate of 3.84 Mbit/s can be
recovered in a cost-efficient way from the line bit rate taking into account the 8B/10B line coding defined in
Section 4.2.5. For example, the 1228.8 Mbit/s correspond to an encoder rate of 122.88 MHz for the 8B/10B
encoder and a subsequent frequency division by a factor of 32 provides the basic UMTS chip rate.
4.2.2. Physical Layer Modes
CPRI is specified for several applications with different interface line bit rates and REC to RE ranges. Table 2
defines several CPRI physical layer modes:

CPRI
CPRI Specification V4.2 (2010-09-29) 29
Table 2: CPRI physical layer modes
Optical Line bit rate Electrical
Short range Long range
614.4 Mbit/s E.6 OS.6 OL.6
1228.8 Mbit/s E.12 OS.12 OL.12
2457.6 Mbit/s E.24 OS.24 OL.24
3072.0 Mbit/s E.30 OS.30 OL.30
4915.2 Mbit/s E.48 OS.48 OL.48
6144.0 Mbit/s E.60 OS.60 OL.60
9830.4 Mbit/s E.96 OS.96 OL.96

For each of those CPRI modes the layer one shall fulfil the requirements as specified in Section 3.5 (clock
stability and noise) and Sections 3.8.2 and 3.8.3 (BER < 10
-12
).
Four electrical variants are recommended for CPRI usage, denoted HV (high voltage), LV (low voltage), LV-II
(low voltage II) and LV-III (low voltage III) in Figure 6A below. The HV variant is guided by IEEE 802.3-2005
[1], clause 39 (1000base-CX) but with 100 ohm impedance. The LV variant is guided by IEEE 802.3-2005 [1]
clause 47 (XAUI) but with lower bit rate. The LV-II variant is guided by OIF-CEI-02.0, clause 7, but with lower
bit rate. The LV-III variant is guided by IEEE 802.3 [22], clause 72.7 and 72.8 (10GBase-KR). See annex 6.2
for more details on the adaptation to CPRI line bit rates and applications.


Figure 6A: HV (high voltage), LV (low voltage), LV-II and LV-III electrical layer 1 usage

It is recommended to reuse optical transceivers from the following High Speed Serial Link standards:

Gigabit Ethernet: Standard IEEE 802.3-2005 [1] clause 38 (1000BASE-SX/LX)
10 Gigabit Ethernet: Standard IEEE 802.3-2005 [1] clause 53 (10GBASE-LX4)
Fibre channel (FC-PI) Standard ISO/IEC 14165-115 [3]
Fibre channel (FC-PI-4) INCITS (ANSI) Revision 8, T11/08-138v1 [18]
Infiniband Volume 2 Rel 1.1 (November 2002) [6]
10 Gigabit Ethernet: Standard IEEE 802.3-2008 [22], Clause 52(10GBASE-S/L/E)
It is recommended to use an optical solution which allows for reuse of SERDES components supporting at
least one of the HV, LV, LV-II, LV-III electrical variants.

The specification does not preclude the usage of any other technique that is proven to reach the same BER
performance (BER < 10
-12
) and clock stability for the dedicated CPRI application.

CPRI clock tolerance is driven by 3GPP requirements (see 3GPP TS 25.104 [8]), which fully permits the
usage of existing high speed serial link standards.

CPRI
CPRI Specification V4.2 (2010-09-29) 30
4.2.3. Electrical Interface
4.2.3.1. Electrical Cabling
No specific cabling is recommended by CPRI.
The cable performance shall be such that transmitter and receiver performance requirements in section 3 are
fulfilled. See also annex 6.2 for explicit recommendations on electrical characteristics.
4.2.3.2. Electrical Connectors
CPRI electrical implementation may use connector solutions that are described and defined in ISO/IEC
14165-115 (Fibre channel FC-PI) [3], INCITS Fibre channel FC-PI-4 [18] or IEEE 802.3-2005 [1].
These solutions are known to achieve the performance required in section 3. See also annex 6.2 for explicit
recommendations on electrical characteristics.
4.2.4. Optical Interface
4.2.4.1. Optical Cabling
The cable performance shall be such that transmitter and receiver performance requirements in section 3 are
fulfilled. The fiber cables recommended for CPRI are:
IEC 60793-2-10:2002.Type A1a (50/125 m multimode) [4]
IEC 60793-2-10:2002.Type A1b (62.5/125 m multimode) [4]
IEC 60793-2-50:2002.Type B1 (10/125 m single-mode) [5]
The exception characteristic as specified in IEEE 802.3-2005 [1] Table 38-12 and IEEE 802.3-2005 [1] Table
53-14 as well as INCITS Fibre channel FC-PI-4 [18] Table 6 and Table 10 may be taken into account.
4.2.4.2. Optical Connectors
CPRI optical implementation may use connector solutions that are described and defined in ISO/IEC 14165-
115 [3] (Fibre channel FC-PI), INCITS Fibre channel FC-PI-4 [18] or IEEE 802.3-2005 [1].
These solutions are known to achieve the performance requirements in section 3. A high flexibility in the
choice of connector and transceiver can be achieved by adopting the SFP [19] and SFP+ [20], [21] building
practice.
4.2.5. Line Coding
8B/10B line coding shall be used for serial transmission according to IEEE 802.3-2005 [1], clause 36.
4.2.6. Bit Error Correction/Detection
The physical layer is designed in such a way that a very low bit error ratio can be achieved without expensive
forward error correction schemes (see requirement R-27). Therefore, no general bit error correction is
applied at layer 1. Some layer 1 control bits have their own protection, see chapter 4.2.7.6.2. The RE and the
REC shall support detection of 8B/10B code violations. Link failures shall be detected by means of 8B/10B
code violations.
4.2.7. Frame Structure
4.2.7.1. Basic Frame Structure
4.2.7.1.1. Framing Nomenclature
The length of a basic frame is 1 T
C
= 1/fc = 1/3.84 MHz = 260.416667ns. A basic frame consists of 16 words
with index W=015. The word with the index W=0, 1/16 of the basic frame, is used for one control word.

CPRI
CPRI Specification V4.2 (2010-09-29) 31
The length T of the word depends on the CPRI line bit rate as shown in Table 3. Each bit within a word is
addressed with the index B, where B=0 is the LSB and B=T-1 is the MSB. Each BYTE within a word is
addressed with the index Y, where B=0 is LSB of Y=0, B=7 is MSB of Y=0, B=8 is LSB of Y=1, etc... For the
notation #Z.X.Y please refer to Section 4.2.7.3.
Table 3: Length of control word
CPRI line bit rate
[Mbit/s]
length of word
[bit]
control word consisting of BYTES with
index
614.4 T=8 Z.X.0
1228.8 T=16 Z.X.0, Z.X.1
2457.6 T=32 Z.X.0, Z.X.1, Z.X.2, Z.X.3
3072.0 T=40 Z.X.0, Z.X.1, Z.X.2, Z.X.3, Z.X.4
4915.2 T=64 Z.X.0, Z.X.1, Z.X.2, Z.X.3, Z.X.4,
Z.X.5, Z.X.6, Z.X.7
6144.0 T=80 Z.X.0, Z.X.1, Z.X.2, Z.X.3, Z.X.4,
Z.X.5, Z.X.6, Z.X.7, Z.X.8, Z.X.9
9830.4 T=128 Z.X.0, Z.X.1, Z.X.2, Z.X.3, Z.X.4, Z.X.5,
Z.X.6, Z.X.7, Z.X.8, Z.X.9, Z.X.10, Z.X.11,
Z.X.12, Z.X.13, Z.X.14, Z.X.15

The remaining words (W=115), 15/16 of the basic frame, are dedicated to the U-plane IQ-data transport
(IQ data block).
4.2.7.1.2. Transmission Sequence and Scrambling
The control BYTES of one basic frame are always transmitted first. The basic frame structure is shown in
Figure 7 to Figure 9A for different CPRI line bit rates. A generic basic frame structure for different line rates is
shown in Figure 9B.
The bit assignment within a BYTE is aligned with IEEE 802.3-2005 [1], namely bit 7 (MSB) = H to bit 0 (LSB)
= A. The physical transmission sequence of the encoded data is defined by the 8B/10B standard according
to IEEE 802.3-2005 [1]. The transmission sequence of the BYTES is indicated on the right hand side of
Figure 7 to Figure 9B with one ball representing a BYTE. After 8B/10B encoding the 10bit code-groups
(abcdei fghj) are transmitted as serial data stream with bit a first.
If the protocol version BYTE #Z.2.0 is set to 2 all data shall be scrambled before 8B/10B line coding by a
side-stream scrambler except for control BYTES #Z.X.Y with index Y1 of subchannel Ns=0 and subchannel
Ns=2. Any seed including zero is allowed (see Annex 6.5 for more details on the scrambling
mechanism).
A device being capable of supporting scrambling (according to annex 6.5) with any seed is defined to be a
device supporting both protocol versions, #Z.2.0=2 and #Z.2.0=1. When transmitting (respectively receiving)
with protocol version #Z.2.0=1 scrambling (respectively descrambling) shall be switched off, which can be
achieved by setting the seed to zero. The protocol version is used in the start-up sequence as specified in
section 4.5.


CPRI
CPRI Specification V4.2 (2010-09-29) 32
1 chip = 1/3.84MHz
1 control word
time
B=0: A
B=1: B
C
D
E
F
G
B=7: H
15 * 8bit
IQ
Data block
B
Y
T
E

#
Z
.
X
.
0
W = 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11,12,13,14,15
Y = 0

Figure 7: Basic frame structure for 614.4 Mbit/s CPRI line bit rate

1 chip = 1/3.84MHz
1 control word 15 * 16 bit
IQ
Data block
B=0: A
B=1: B
C
D
E
F
G
H
A
B
C
D
E
F
G
B=15: H
B
Y
T
E

#
Z
.
X
.
0
B
Y
T
E


#
Z
.
X
.
1
time
W = 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11,12,13,14,15
Y = 0
Y = 1

Figure 8: Basic frame structure for 1228.8 Mbit/s CPRI line bit rate

CPRI
CPRI Specification V4.2 (2010-09-29) 33
1 chip = 1/3.84MHz
1 control word
15 * 32 bit
IQ
Data block
B
Y
T
E

#
Z
.
X
.
2
B
Y
T
E


#
Z
.
X
.
3
B=0: A
B=1: B
C
D
E
F
G
H
A
B
C
D
E
F
G
H
A
B
C
D
E
F
G
H
A
B
C
D
E
F
G
B=31: H
B
Y
T
E

#
Z
.
X
.
0
B
Y
T
E


#
Z
.
X
.
1
time
W = 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11,12,13,14,15
Y = 0
Y = 1
Y = 2
Y = 3


Figure 9: Basic frame structure for 2457.6 Mbit/s CPRI line bit rate

CPRI
CPRI Specification V4.2 (2010-09-29) 34
1 chip = 1/3.84MHz
IQ
Data block
B
Y
T
E

#
Z
.
X
.
2
B
Y
T
E


#
Z
.
X
.
3
B=0: A
B=1: B
C
D
E
F
G
H
A
B
C
D
E
F
G
H
A
B
C
D
E
F
G
H
A
B
C
D
E
F
G
H
B
Y
T
E

#
Z
.
X
.
0
B
Y
T
E


#
Z
.
X
.
1
time
W = 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11,12,13,14,15
Y = 0
Y = 1
Y = 2
Y = 3
1 control word
15 * 40 bit
B
Y
T
E

#
Z
.
X
.
4
A
B
C
D
E
F
G
B=39: H
Y = 4

Figure 9A: Basic frame structure for 3072.0 Mbit/s CPRI line bit rate

CPRI
CPRI Specification V4.2 (2010-09-29) 35
B
Y
T
E

#
Z
.
X
.
0
B
Y
T
E


#
Z
.
X
.
1
B
Y
T
E

#
Z
.
X
.
(
T
/
8
-
1
)

Figure 9B: Generic basic frame structure for different CPRI line rates (T is defined in Table 3)

4.2.7.2. Mapping of IQ-data
4.2.7.2.1. IQ Sample Widths
The required sample width of the user-plane IQ-data depends on the application layer. This specification
provides a universal mapping scheme in order to implement any of the required sample widths depending on
the application layer. The option list for I and Q samples can be found in Table 4. Mixed sample widths within
one basic frame are not described in detail but are allowed if required. One IQ sample consists of one I
sample and one equal-sized Q sample.
Table 4: Option list for I and Q sample width ranges
Direction of link
Symbol for sample
width
Range
[bits]
Downlink M 8, 9, 10, , 20
Uplink M 4, 5, 6, , 20

4.2.7.2.2. Mapping of IQ Samples within one AxC Container
An AxC Container is a sub-part of the IQ-data block of a basic frame.
For UTRA-FDD, an AxC Container contains exactly n IQ samples from the same AxC, where n is
the oversampling ratio with respect to the chip rate f
C
= 3.84MHz. The oversampling ratio n is
defined in Table 5 and Table 5A. For UTRA-FDD the sampling rate is given by f
S
=nf
C
.

CPRI
CPRI Specification V4.2 (2010-09-29) 36
For WiMAX, an AxC Container contains IQ sample bits and/or stuffing bits. One of the IQ mapping
methods 1, 2 or 3, as specified in the following sections, shall apply per WiMAX AxC. For WiMAX the
sampling rate f
S
can be derived from the definitions given in [11].
For E-UTRA, an AxC Container contains IQ sample bits from the same AxC and/or stuffing bits.
The E-UTRA IQ-samples shall be mapped to the AxC Container according to Mapping method 1
(section 4.2.7.2.5) or Mapping method 3 (section 4.2.7.2.7). For E-UTRA the typical sampling rates
f
S
can be derived from the 3GPP TS 36.104 [14] and 36.211 [16] as described in Annex 6.4.
The size of one AxC Container N
AxC
shall be an even number of bits.
IQ sample(s) shall be sent in an AxC Container in the following way:
from LSB (I
0
, Q
0
) to MSB (I
M-1
, Q
M-1
) or (I
M-1
, Q
M-1
),
I and Q samples being interleaved,
in chronological order and consecutively.
The option lists for uplink and downlink oversampling ratios n can be found in Table 5 and Table 5A,
respectively. The oversampling ratios of uplink and downlink may be selected independently.
Table 5: Option list for UTRA FDD UL oversampling ratios n with respect to f
C

Opt. 1 Opt. 2
UL Oversampling Ratio n 2 4
UL Symbols for IQ samples I, Q, I, Q I, Q, I, Q, I, Q, I, Q

Table 5A: Option list for UTRA FDD DL oversampling ratios n with respect to f
C

Opt. 1 Opt. 2
DL Oversampling Ratio n 1 2
DL Symbols for IQ samples I, Q I, Q, I, Q

The IQ sample widths and the oversampling ratios for downlink and uplink shall be decided on application
layer per AxC. Figure 10 to Figure 12 show the IQ sample arrangement and the transmission order for uplink
and downlink for the described oversampling options.
Q
2
Q
1
Q
0
I
2
I
1
I
0
Q
M-1
Q
M-2
I
M-1
I
M-2
Q
2
Q
1
Q
0
I
2
I
1
I
0
Q
M-1
Q
M-2
I
M-1
I
M-2

Figure 10: IQ samples within one AxC with oversampling ratio 1

Q
2
Q
1
Q
0
I
2
I
1
I
0
Q
M-1
Q
M-2
I
M-1
I
M-2
Q
2
Q
1
Q
0
I
2
I
1
I
0
Q
M-1
Q
M-2
I
M-1
I
M-2
Q
2
Q
1
Q
0
I
2
I
1
I
0
Q
M-1
Q
M-2
I
M-1
I
M-2
Q
2
Q
1
Q
0
I
2
I
1
I
0
Q
M-1
Q
M-2
I
M-1
I
M-2
Q
2
Q
1
Q
0
I
2
I
1
I
0
Q
M-1
Q
M-2
I
M-1
I
M-2
Q
2
Q
1
Q
0
I
2
I
1
I
0
Q
M-1
Q
M-2
I
M-1
I
M-2

Figure 11: IQ samples within one AxC with oversampling ratio 2 (uplink direction shown; for the downlink
direction M shall be replaced by M)


CPRI
CPRI Specification V4.2 (2010-09-29) 37
Q
2
Q
1
Q
0
I
2
I
1
I
0
Q
M-
1
Q
M-
2
I
M-1
I
M-2
Q
2
Q
1
Q
0
I
2
I
1
I
0
Q
M
-1
Q
M
-2
I
M-1
I
M-2
Q
2
Q
1
Q
0
I
2
I
1
I
0
Q
M-1
Q
M-2
I
M-1
I
M-2
Q
2
Q
1
Q
0
I
2
I
1
I
0
Q
M-
1
Q
M-2
I
M-1
I
M-2
Q
2
Q
1
Q
0
I
2
I
1
I
0
Q
M-
1
Q
M-
2
I
M-1
I
M-2
Q
2
Q
1
Q
0
I
2
I
1
I
0
Q
M
-1
Q
M
-2
I
M-1
I
M-2
Q
2
Q
1
Q
0
I
2
I
1
I
0
Q
M-1
Q
M-2
I
M-1
I
M-2
Q
2
Q
1
Q
0
I
2
I
1
I
0
Q
M-
1
Q
M-2
I
M-1
I
M-2

Figure 12: IQ samples within one uplink AxC with oversampling ratio 4
4.2.7.2.3. Mapping of AxC Container within one Basic Frame
The following mapping rules apply for both, uplink and downlink:
Each AxC Container is sent as a block.
Overlap of AxC Containers is not allowed.
The position of each AxC Container in the IQ data block is decided by one of the following options:
o Option 1 (packed position):
Each AxC Container in a basic frame is sent consecutively (without any reserved bits in
between) and in ascending order of AxC number.
o Option 2 (flexible position):
For each AxC Container, the application shall decide at what address (W, B for W>0) in
the IQ data block the first bit of the AxC Container is positioned. The first bit of an AxC
Container shall be positioned on an even bit position in the IQ data block (B shall be even).
The bits not used by AxC Containers in the IQ data block in the basic frame shall be treated as
reserved bits (r).
Figure 13 illustrates these mapping rules for both mapping options.

Figure 13: Example of AxC Container mapping in the IQ data block
4.2.7.2.4. Common properties of IQ mapping methods
Transmission of WiMAX/E-UTRA AxCs is organized in a consecutive flow of AxC Container Blocks, where
each AxC Container Block has the duration of K basic frames. There are S IQ samples per WiMAX/E-
UTRA AxC being carried in one AxC Container Block. The S IQ samples per WiMAX/E-UTRA AxC are
mapped into the AxC Container Block in chronological order as shown in Figure 13A. Consecutive AxC
Container Blocks construct a bit pipe. IQ samples with stuffing bits are arranged into the pipe as a
continuous bit sequence. The synchronization between AxC Container Blocks and CPRI framing is
specified in section 4.2.8.

CPRI
CPRI Specification V4.2 (2010-09-29) 38

Figure 13A: Relation between S IQ samples and one AxC Container Block
S and K are nonzero integers. Different mapping methods provide different definitions for S and K as
described in the sections 4.2.7.2.5, 4.2.7.2.6, and 4.2.7.2.7. For each AxC, the mapping method and the
associated parameters (e.g. S, K values) are decided by the application layer in the REC
5
. The information is
then sent to the RE(s) through the C&M channel.
4.2.7.2.5. Mapping method 1: IQ sample based
This mapping method is intended for dense packing of IQ data into the CPRI data flow (high bandwidth
efficiency) and is optimized for low latency together with sample based processing of IQ data in the RE(s).
For this mapping method the size N
AxC
of the AxC Container shall be chosen according to equation (3).

=
C
S
AxC
f
f M
ceil 2 N
(3)
The function ceil returns the smallest integer greater than or equal to the argument.
M is the width of I or Q sample for downlink as defined in Table 4. M shall be used instead of M for the uplink
case. If no further information is given, the same rules shall be used for both, downlink and uplink.
For this mapping method the S and K shall satisfy equation (4).
C S
f f
K S
= (4)
S and K shall be calculated using equations (5) and (6).
( )
S
C S
f
f , f LCM
= K , (5)
( )
C
C S
f
f , f LCM
= S , (6)
where LCM means Least Common Multiple.
For this mapping method one AxC Container Block contains two parts, as shown in Figure 13B: The first
part is filled with a number N
ST
= KN
AxC
2MS of stuffing bits; the second part is filled with S samples. The
stuffing bits shall be vendor specific (v).


5
An RE may not support all mapping methods. The REC shall take the capabilities of the RE into consideration for its decision.
one AxC Container Block (duration: K basic frames)
0 1 2 K-2 K-1
S IQ samples (stuffing bits not shown)
0 1 2 S-3 S-2 S-1
t

CPRI
CPRI Specification V4.2 (2010-09-29) 39

Figure 13B: IQ Sample based mapping in an AxC Container Block
4.2.7.2.6. Mapping method 2: WiMAX symbol based
This mapping method is intended for dense packing of IQ data into the CPRI data flow and is optimized for
low latency together with WiMAX symbol based processing of IQ data in the RE(s).
The length K of the AxC Container Block shall be chosen equal to the WiMAX frame duration T
F
, as
described by the following equation (7).
C F
f T = K (7)
For all WiMAX frame durations T
F
defined in [11], K is an integer. The AxC Container Block shall be aligned
with the WiMAX frame.
For this mapping method one AxC Container Block contains two parts: The first part is filled with N
SYM
AxC
Symbol Blocks; the second part is filled with N
S_FRM
stuffing bits
6
. N
SYM
is the number of WiMAX symbols in
one WiMAX frame as given by equation (8), where T
S
is the duration of one symbol as defined in [13] section
8.3.2.2.

=
S
F
SYM
T
T
floor N (8)
The function floor returns the greatest integer less than or equal to the argument.
In each AxC Symbol Block, there are also two parts: The first part is filled with N
SAM
samples; the second
part is filled with N
S_SYM
stuffing bits. N
SAM
is the number of samples (either with or without CP) during one
WiMAX symbol.
The total number of S samples per AxC Container Block is given by equation (9):
SAM SYM
N N S =

(9).
All of these relations are illustrated in Figure 13C.

6
The N
S_FRM
stuffing bits are required since the length of a WiMAX frame is in general not an integer multiple of symbol lengths.

t
#
one AxC Container Block
N
ST
stuffing bits
#+1 #-1
0
bits of one sample
bits of S samples
1 2 S-1 S-2

CPRI
CPRI Specification V4.2 (2010-09-29) 40

Figure 13C: Symbol based mapping in an AxC Container Block
For this mapping method the size N
AxC
of the AxC Container shall be chosen according to inequality (10).

K
S
N
M
ceil 2
AxC
(10)
The number N
S_SYM
of stuffing bits in one AxC Symbol Block and the number N
S_FRM
of stuffing bits in one
AxC Container Block are given by equations (11) and (12), respectively.

=
SYM
AxC
S_SYM
M 2
floor
N
S N K
N
(11)
SYM S_SYM AxC S_FRM
M 2 N N S N K N = (12)
4.2.7.2.7. Mapping method 3: Backward compatible
For this mapping method the size of the AxC Container N
AxC
= 2M shall be chosen with M being in the
range as specified in Table 4.
This choice makes use of the AxC Containers which have been defined for UTRA-FDD in CPRI releases 1
and 2 for downlink. For uplink the same mapping method shall apply as for downlink. WiMAX/E-UTRA can
be implemented as an application above a CPRI release 1 or 2 communication as shown in Figure 13D. One
AxC Container contains exactly one sample (which could be a stuffing sample in case of WiMAX/E-UTRA).
With this mapping method WiMAX/E-UTRA can easily be implemented in networking topologies where CPRI
release 1 or 2 compatible REs already exist.

t
N
S_FRM
stuffing bits
0 1
bits of one WiMAX sample
bits of N
SAM
WiMAX samples
N
SAM
-1
bits of one AxC Symbol Block
N
S_SYM
stuffing bits
bits of N
SYM
AxC Symbol Blocks
N
SYM
-2 N
SYM
-1

0 1
#+1 # #-1
one AxC Container Block

CPRI
CPRI Specification V4.2 (2010-09-29) 41

Figure 13D: Example of protocol stack based upon CPRI release 1 and 2

For this mapping method S and K shall be calculated by equations (5) and (6) as with IQ sample based
mapping in section 4.2.7.2.5.
Multiplexed IQ samples of an AxC Group are carried in AxC Container Groups consisting of N
C
AxC
Containers per basic frame.

The AxC Group contains N
A
AxCs (AxC#0, AxC#1, , AxC# N
A
-1). However, it is not mandatory to
handle AxCs with same features in an AxC Group, therefore N
A
=1 is the basic configuration.

One AxC Container Block contains N
A
S samples.

N
C
shall satisfy inequality (13).

K
S N
N
A
C
ceil (13)

N
C
should be chosen by equation (14) in order to minimize the number of stuffing samples N
V
that is defined
in equation (15).

=
K
S N
N
A
C
ceil (14)

Within one AxC Group all samples shall have the same width M and all AxC Containers shall have the
same size N
AxC
= 2*M (Each IQ sample is stored in an AxC Container as specified in CPRI release 1 and 2).

One AxC Container Block contains N
C
K AxC Containers, which are indexed in chronological order from
k=0 to k=N
C
*K-1. The number N
V
of stuffing samples per AxC Container Block is given by the equation
(15):
S N K N N =
A C V
(15)
For WiMAX, the values for S and K, as well as the recommended values for N
C
and N
V
are provided in Table
5B for the basic configuration with N
A
=1 and N
A
=2 and for the sampling rates f
S
as specified in [11]. For E-
UTRA, the corresponding values for the sampling rates listed in Annex 6.4 are shown in Table 5C.


RE
Layer 1
Layer 2
Control &
Mgmt
User Plane Sync
SAP
CM
Layer 1
Layer 2
Control &
Mgmt
User Plane Sync
Radio Base Station System
CPRI link
SAP
S
SAP
IQ
SAP
CM
SAP
S
SAP
IQ
Master port Slave port
REC
AxC Group
MUX/DEMUX
Application
(WiMAX, E-UTRA
or UMTS)
IQ samples
AxC #0 #1 #2 Stuffing
Samples
v
Stuffing
Samples
v
IQ samples
AxC #0 #1 #2
Network
interface
Air
interface
MUX/DEMUX
SAP
IQ
SAP
IQ
AxC Group
CPRI
Release 1 or 2

CPRI
CPRI Specification V4.2 (2010-09-29) 42
Table 5B: Recommended number N
V
of stuffing samples for N
A
=1 and N
A
=2 (WiMAX)
f
S
[MHz] N
A
S K N
C
N
V
= N
C
K- N
A
S
4 1 25 24 2 23
5.6 1 35 24 2 13
8 1 25 12 3 11
10 1 125 48 3 19
11.2 1 35 12 3 1
4 2 25 24 3 22
5.6 2 35 24 3 2
8 2 25 12 5 10
10 2 125 48 6 38
11.2 2 35 12 6 2

Table 5C: Recommended number N
V
of stuffing samples for N
A
=1 and N
A
=2 (E-UTRA)
f
S
[MHz] N
A
S K N
C
N
V
= N
C
K- N
A
S
1.92

1 1 2 1 1
3.84 1 1 1 1 0
7.68 1 2 1 2 0
15.36 1 4 1 4 0
23.04 1 6 1 6 0
30.72 1 8 1 8 0
1.92

2 1 2 1 0

In case of N
V
>0 the position k
i
of each stuffing sample i within the k=0 to k=N
C
*K-1 AxC Containers is given
by

=
V
C
i
floor
N
K N i
k ; for i=0,1,, N
V
-1 (16)
The AxC Containers with index k
i
are filled with stuffing samples which consist of vendor specific bits v. All
remaining AxC Containers in the AxC Container Block are filled with samples of AxC#0, AxC#1,
AxC#2,, AxC#N
A
-1 in chronological order. This mapping method is illustrated in Figure 13E.

CPRI
CPRI Specification V4.2 (2010-09-29) 43
Figure 13E: Example of an AxC Group with N
A
=2 (AxC#0, AxC#1) mapped into an AxC Container Group
with N
C
=6 AxC Containers per basic frame (AxC Container #0 through AxC Container #5)
4.2.7.2.8. WiMAX/E-UTRA TDD and WiMAX/E-UTRA FDD
Both TDD and FDD have the same AxC Container definition and mapping rules as in the former sections.
During the TDD sub-frame for uplink, there will be no IQ sample transfer in downlink, and the transmitter
shall send stuffing bits v. During the TDD sub-frame for downlink, there will be no IQ sample transfer in
uplink, and the transmitter shall send stuffing bits v.
TDD switching points in each WiMAX/E-UTRA frame shall be defined by the application layer in the REC,
and be sent through the C&M channel to the RE(s).

4.2.7.3. Hyperframe Structure
The hyperframe structure is hierarchically embedded between the basic frame and the CRPI 10ms frame as
shown in Figure 14.


CPRI
CPRI Specification V4.2 (2010-09-29) 44

BFN
#0 #Z #149 CPRI 10ms frame
(150 hyper frames = 10ms)
#0 #X #255
hyperframe
(256 basic frames = 66.67s)
W
Z: hyperframe number
X: basic frame number
basic frame
(1 Tchip = 260.42ns)
1 15 bytes
Y
8 bits
W: word number in basic
frame

Y: byte number within word

Figure 14: Illustration of the frame hierarchy and notation indices

Z is the hyperframe number, X is the basic frame number within a hyperframe, W is the word number within
a basic frame and Y is the byte number within a word. The control word is defined as word with rank W=0.
The value ranges of the indices are shown in Table 6:
Table 6: Value ranges of indices
CPRI line bit
rate
[Mbit/s]
Z X W Y B
614.4 0 0, 1, 7
1228.8 0, 1 0, 1, 15
2457.6 0, 1, 2, 3 0, 1, 31
3072.0 0, 1, 2, 3, 4 0, 1, 39
4915.2 0, 1, 2, , 7 0, 1, , 63
6144.0 0, 1, 2, , 9 0, 1, , 79
9830.4

0, 1, ..., 149

0, 1, , 255

0, 1, , 15
0, 1, 2, , 15 0, 1, , 127

4.2.7.4. Subchannel Definition
The 256 control words of a hyperframe are organized into 64 subchannels of 4 control words each. One
subchannel contains 4 control words per hyperframe.

The index Ns of the subchannel ranges from 0 to 63. The index Xs of a control word within a subchannel has
four possible values, namely 0, 1, 2 and 3. The index X of the control word within a hyperframe is given by X
= Ns + 64*Xs.

CPRI
CPRI Specification V4.2 (2010-09-29) 45

The organization of the control words in subchannels is illustrated in Figure 15 and Figure 16.

Comma Byte
Synchronization and timing
C&M CPU-CPU link
layer 1 inband protocol
Reserved
Vendor specific
fast C&M
pointer to start of fast C&M
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
* -->
61
62
63
Xs= 0 1 2
0 64
1 65
3 67
4
14
15 79 143 207
16 80 144 208
17
62 126 190 254
61
63 127 191 255
2 66 *
index X of control word
within hyperframe:
X = Ns + 64* Xs
(some indices X are inserted
as examples)
SlowC&M link
Fast C&M link
p
p
Pointer p
Ns=0
Comma Byte
Synchronization and timing
C&M CPU-CPU link
L1 inband protocol
Reserved
Vendor specific
fast C&M
pointer to start of fast C&M
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
* -->
61
62
63
Xs= 0 1 2
0 64
1 65
3 67
4
14
15 79 143 207
16 80 144 208
17
62 126 190 254
61
63 127 191 255
2 66 *
index X of control word
within hyperframe:
X = Ns + 64* Xs
(some indices X are inserted
as examples)
SlowC&M link
Fast C&M link
p
Pointer p
Ns=0

Figure 15: Illustration of subchannels within one hyperframe
1 hyperframe
1 basic frame
index of
control word X=0 1 2 3 15 16 p-1 p 63 64 65 66 67 127 255
index of
subchannel Ns=0 1 2 3 15 16 p-1 p 63 0 2 3 1 63 63
index of
control word
within subchannel
Xs=0 0 0 0 0 0 0 0 0 1 1 1 1 1 3

Figure 16: Illustration of control words and subchannels within one hyperframe

CPRI
CPRI Specification V4.2 (2010-09-29) 46
Table 7: Implementation of control words within one hyperframe for pointer p > 19
subchannel
number Ns
purpose of
subchannel
Xs=0 Xs=1 Xs=2 Xs=3
0 sync&timing sync byte K28.5 HFN BFN-low BFN-high
1 slow C&M slow C&M slow C&M slow C&M slow C&M
2 L1 inband prot. version startup L1-reset-LOS... pointer p
3 reserved reserved reserved reserved reserved
... ... ... ... ... ...
15 reserved reserved reserved reserved reserved
16 vendor specific vendor specific vendor specific vendor specific vendor specific
... ... ... ... ... ...
p-1 vendor specific vendor specific vendor specific vendor specific vendor specific
pointer: p fast C&M fast C&M fast C&M fast C&M fast C&M
... ... ... ... ... ...
63 fast C&M fast C&M fast C&M fast C&M fast C&M


For subchannel 0 the content of the control BYTES #Z.X.Y with index Y1 is reserved (r), except for the
synchronization control word (Xs=0) where Table 9 applies. For subchannel 1 Table 11 applies. For
subchannel 2 the content of the control BYTES #Z.X.Y with index Y1 is reserved (r).
4.2.7.5. Synchronization Data
The following control words listed in Table 8 are dedicated to layer 1 synchronization and timing. The support
of the control words in Table 8 and Table 9 is mandatory.

Table 8: Control words for layer 1 synchronization and timing
BYTE index Function content comment
Z.0.0 Start of hyperframe Special code K28.5
Z.64.0 HFN (Hyperframe
number)
HFN=0149,
the first hyperframe in an
UMTS radio frame has
HFN=0. The exact HFN
bit mapping is indicated in
Figure 17.
Z.128.0
and
Z.192.0
BFN
(CPRI 10ms frame
number; for UTRA FDD
aligned with NodeB
Frame Number)


#Z.128.0 (low byte) and

b3-b0 of #Z.192.0 are
BFN

b7-b4 of #Z.192.0 are
reserved (all r). The
exact mapping is
described in Figure 18.
CPRI 10ms frame
synchronisation, HFN and
BFN are described in
detail in sections 4.2.8
and 4.2.9.

HFN is mapped within #Z.64.0 as defined in Figure 17.

CPRI
CPRI Specification V4.2 (2010-09-29) 47

B7 b0
#Z.64.0

MSB HFN LSB
Figure 17: HFN mapping

BFN is mapped within #Z.128.0 and #Z.192.0 as defined in Figure 18. #Z.192.0 b7---b4 are reserved bits.
B3 b0 b7 b0
#Z.192.0 #Z.128.0

MSB BFN LSB
Figure 18: BFN mapping

Table 9: Synchronization control word
CPRI line bit rate[Mbit/s] 614.4 1228.8 2457.6 3072.0 4915.2 6144.0 9830.4
#Z.0.0
Sync.
Byte
K28.5
(BCh)
K28.5
(BCh)
K28.5
(BCh)
K28.5
(BCh)
K28.5
(BCh)
K28.5
(BCh)
K28.5
(BCh)
D16.2
(50h)
D16.2
(50h)
D16.2
(50h)
D16.2
(50h)
D16.2
(50h)
D16.2
(50h)
#Z.0.1
D5.6
(C5h)
D5.6
(C5h)
D5.6
(C5h)
D5.6
(C5h)
D5.6
(C5h)
D5.6
(C5h)
#Z.0.2
#Z.0.3
D16.2
(50h)
#Z.0.4
#Z.0.5
D16.2
(50h)
#Z.0.6
#Z.0.7
D16.2
(50h)
#Z.0.8
#Z.0.9
D16.2
(50h)

#Z.0.10
#Z.0.11
#Z.0.12
#Z.0.13
#Z.0.14
Sync.
Control
Word
#Z.0.15
Filling
Bytes
N/A
N/A
N/A
N/A
N/A
N/A
D16.2
(50h)
Remark:
The sequences K28.5+D5.6 and K28.5+D16.2 are defined in the 8B/10B standard as /I1/ and /I2/
ordered_sets (IDLE1 sequences with opposing disparity and IDLE2 sequences with preserving disparity) and
are expected to be supported by commonly used SERDES devices.

CPRI
CPRI Specification V4.2 (2010-09-29) 48
According to Table 9, the transmitter may send either D16.2 or D5.6 as BYTE #Z.0.1. The receiver shall
accept both D16.2 and D5.6.
4.2.7.6. L1 Inband Protocol
Reserved bits in this section are marked with r. This means that a transmitter shall send 0s for bits marked
with r, and the receiver shall not interpret bits marked with r (transmit: r = 0, receiver: r = dont care).
The control BYTES listed in Table 10 are dedicated to L1 inband protocol.
Table 10: Control BYTES for L1 inband protocol
BYTE index function content comment
Z.2.0 Protocol version 0000 0001 or 0000 0010 This document refers to
protocol version 1 and 2
Z.66.0 Start-up rrrr rCCC

b2-b0 HDLC bit rate:
000: no HDLC
001: 240kbit/s HDLC
010: 480kbit/s HDLC
011: 960kbit/s HDLC
(for line rates 1228.8Mbit/s)
100: 1920kbit/s HDLC
(for line rates 2457.6Mbit/s)
101: 2400kbit/s HDLC
(for line rates 3072.0Mbit/s)
110: Highest possible HDLC bit
rate
(for line rates > 3072.0Mbit/s)
111: HDLC bit rate negotiated on
higher layer, see section 4.5.3.4.
For an overview refer to Table 11

b7-b3: reserved (all r)
Enables the HDLC link to
be established
Z.130.0 L1 SDI, RAI,
Reset, LOS, LOF


rrrF LSAR

b0: Reset
0: no reset
1: reset
DL: reset request
UL: reset acknowledge

b1: RAI
b2: SDI
b3: LOS
b4: LOF
Basic layer 1 functions

CPRI
CPRI Specification V4.2 (2010-09-29) 49
0: alarm cleared
1: alarm set

b7-b5: reserved (all r)
Z.194.0 Pointer p rrPPPP PP

b5-b0: Pointer to subchannel
number, where Ethernet link
starts:
000000: p=0: no Ethernet
channel
000001

010011: p=119 invalid (no
Ethernet channel, not possible
since other control words would
be affected)
010100:

111111: p=2063: valid
Ethernet channel, for bit rates
refer to Table 12

b7-b6: reserved (all r)
Indicates the subchannel
number Ns at which the
control words for the
Ethernet channel starts
within a hyperframe.


4.2.7.6.1. Reset
Reset of the link is managed through start-up sequence definition (see Section 4.5). Reset of the RE is
managed with the Reset bit in #Z.130.0. The reset notification can only be sent from a master port to a slave
port. The reset acknowledgement can only be sent from a slave port to a master port. When the master
wants to reset a slave, it shall set DL #Z.130.0 b0 for at least 10 hyperframes. On the reception of a valid
reset notification, the slave shall set UL #Z.130.0 b0 at least 5 hyperframes on the same link.
When an RE receives a valid reset notification on any of its slave ports, it shall not only reset itself, but also
forward reset notification on all its master ports by setting DL #Z.130.0 b0 for at least 10 hyperframes.
While in reset and if the link is still transmitting, the RE must set the SDI bit.
4.2.7.6.2. Protection of Signalling Bits
Signalling bits shall be protected by filtering over multiple hyperframes. The filtering shall be done by a
majority decision of the 5 instances of one signalling bit derived from the 5 most recent hyperframes. The
filtering guarantees that 2 consecutive erroneous receptions of instances of one signalling bit do not result in
an erroneous interpretation.
This filtering requirement applies to the following signalling bit:
#Z.130.0, b0: R (Reset) in both DL and UL.
The filtering of the other inband protocol bits, i.e., #Z.66.0 (HDLC rate), #Z.194.0 (pointer to Ethernet
channel), #Z.130.0 (layer 1 link maintenance) and #Z.2.0 (protocol version) shall be performed by the
application layer (see also Section 4.2.10).


CPRI
CPRI Specification V4.2 (2010-09-29) 50
4.2.7.7. C&M Plane Data Channels
CPRI supports two different types of C&M channels, which shall be selected from the following option list:
C&M Channel Option 1: Slow C&M Channel based on HDLC
C&M Channel Option 2: Fast C&M Channel based on Ethernet
4.2.7.7.1. Slow C&M Channel
One option is to use a low rate HDLC channel for C&M data. The bit rate is defined by the 3 LSBs of the
start-up information BYTE #Z.66.0 (see Table 11). The mapping of control BYTES to HDLC serial data is
according to what is shown for the different configurations in Figure 19 to Figure 22B.
Parameter T used in Table 11 is defined in Table 3.

CPRI
CPRI Specification V4.2 (2010-09-29) 51
Table 11: Achievable HDLC bit rates in kbit/s
CPRI line
bit rate
[Mbit/s]
#Z.66.0=
rrrr r000
#Z.66.0=
rrrr r001
#Z.66.0=
rrrr r010

#Z.66.0=
rrrr r011
#Z.66.0=
rrrr r100
#Z.66.0=
rrrr r101
#Z.66.0=
rrrr r110
#Z.66.=
rrrr r111
614.4 no HDLC 240 480 invalid invalid invalid invalid
1228.8 no HDLC 240 480 960 invalid invalid invalid
2457.6 no HDLC 240 480 960 1920 invalid invalid
3072.0 no HDLC 240 480 960 1920 2400 invalid
4915.2 no HDLC 240 480 960 1920 2400 3840
6144.0 no HDLC 240 480 960 1920 2400 4800
9830.4 no HDLC 240 480 960 1920 2400 7680
used
control
BYTE
indices for
the HDLC
channel
and their
sequential
order
no HDLC Z.1.0
Z.129.0
Z.1.0
Z.65.0
Z.129.0
Z.193.0
Z.1.0
Z.1.1
Z.65.0
Z.65.1
Z.129.0
Z.129.1
Z.193.0
Z.193.1
Z.1.0
Z.1.1
Z.1.2
Z.1.3
Z.65.0
Z.65.1
Z.65.2
Z.65.3
Z.129.0
Z.129.1
Z.129.2
Z.129.3
Z.193.0
Z.193.1
Z.193.2
Z.193.3
Z.1.0
Z.1.1
Z.1.2
Z.1.3
Z.1.4
Z.65.0
Z.65.1
Z.65.2
Z.65.3
Z.65.4
Z.129.0
Z.129.1
Z.129.2
Z.129.3
Z.129.4
Z.193.0
Z.193.1
Z.193.2
Z.193.3
Z.193.4
Z.1.0
Z.1.1
.
Z.1.(T/8-1)
Z.65.0
Z.65.1
..
Z.65.(T/8-1)
Z.129.0
Z.129.1
.
Z.129.(T/8-1)
Z.193.0
Z.193.1
.
Z.193.(T/8-1)
See
section
4.5.3.4
and
section
4.5.3.5.
Remark: In case of an invalid configuration no HDLC shall be used.


CPRI
CPRI Specification V4.2 (2010-09-29) 52
#Z.1.0
01111110 Address FCS 01111110 FCS
HDLC-Frame n-1 HDLC-Frame n HDLC-Frame n+1
01111110 01111110 01111110
#Z.129.0 #Z+1.1.0 #Z+1.129.0 #Z+2.1.0
time
bit 0(LSB) bit 7(MSB)

Figure 19: Mapping of control BYTES to HDLC serial data with 240kbit/s
#Z.1.0
01111110 Address FCS 01111110 FCS
HDLC-Frame n-1 HDLC-Frame n HDLC-Frame n+1
01111110 01111110 01111110
#Z.65.0 #Z.129.0 #Z.193.0 #Z+1.1.0
time
bit 0(LSB) bit 7(MSB)

Figure 20: Mapping of control BYTES to HDLC serial data with 480kbit/s
#Z.1.0
01111110 Address FCS 01111110 FCS
HDLC-Frame n-1 HDLC-Frame n HDLC-Frame n+1
01111110 01111110 01111110
#Z.1.1 #Z.65.0 #Z.65.1 #Z.129.0
time
bit 0(LSB) bit 7(MSB)

Figure 21: Mapping of control BYTES to HDLC serial data with 960kbit/s

CPRI
CPRI Specification V4.2 (2010-09-29) 53
#Z.1.0
01111110 Address FCS 01111110 FCS
HDLC-Frame n-1 HDLC-Frame n HDLC-Frame n+1
01111110 01111110 01111110
#Z.1.1 #Z.1.2 #Z.1.3 #Z.65.0
time
bit 0(LSB) bit 7(MSB)

Figure 22: Mapping of control BYTES to HDLC serial data with 1920kbit/s
#Z.1.0
01111110 Address FCS 01111110 FCS
HDLC-Frame n-1 HDLC-Frame n HDLC-Frame n+1
01111110 01111110 01111110
#Z.1.1 #Z.1.2 #Z.1.3 #Z.1.4
time
bit 0(LSB) bit 7(MSB)
#Z.65.0

Figure 22A: Mapping of control BYTES to HDLC serial data with 2400kbit/s



Figure 22B: Mapping of control BYTES to HDLC serial data for #Z.66.0 = rrrr r110 (T is defined in Table 3)

4.2.7.7.2. Fast C&M Channel
Another option is to use a high data rate Ethernet Channel which can be flexibly configured by the pointer in
control BYTE #Z.194.0. The mapping of the Ethernet data follows the same principle as the HDLC channel
(no byte alignment, LSB first).

CPRI
CPRI Specification V4.2 (2010-09-29) 54
The Ethernet bit rate is configured with the pointer in control BYTE #Z.194.0. In contrast to the HDLC link,
the full control words shall always be used for the Ethernet channel. The achievable Ethernet bit rates are
shown in Table 12.
Table 12: Achievable Ethernet bit rates
CPRI line
bit rate
[Mbit/s]
length of
control
word [bit]
control word consisting
of BYTES with index
minimum Ethernet bit
rate [Mbit/s]
(#Z.194.0=rr111111)
maximum Ethernet bit
rate [Mbit/s]
(#Z.194.0=rr010100)
614.4 8 Z.X.0 0.48 21.12
1228.8 16 Z.X.0, Z.X.1 0.96 42.24
2457.6 32 Z.X.0, Z.X.1, Z.X.2, Z.X.3 1.92 84.48
3072.0 40 Z.X.0, Z.X.1, Z.X.2, Z.X.3,
Z.X.4
2.4 105.6
4915.2 64 Z.X.0, Z.X.1, Z.X.2, Z.X.3,
Z.X.4, Z.X.5, Z.X.6, Z.X.7
3.84 168.96
6144.0 80 Z.X.0, Z.X.1, Z.X.2, Z.X.3,
Z.X.4, Z.X.5, Z.X.6, Z.X.7,
Z.X.8, Z.X.9
4.8 211.2
9830.4 128 Z.X.0, Z.X.1, Z.X.2, Z.X.3,
Z.X.4, Z.X.5, Z.X.6, Z.X.7,
Z.X.8, Z.X.9, Z.X.10, Z.X.11,
Z.X.12, Z.X.13, Z.X.14, Z.X.15
7.68 337.92

Packet detection, start and termination is based on SSD and ESD coding sequence as shown in Figure 23.
#Z.63.0
SSD 10bit Ethernet packet IDLE 10bit
4B/5B encoded data from Ethernet MAC (LSB first)
01111110 ESD 10bit
#Z.63.1
time
bit 0(LSB)
#Z.127.0 #Z.127.1 #Z.191.1 #Z.191.0
#Z.255.0 #Z+1.63.1 #Z+1.63.0 #Z.255.1

Figure 23: Example showing the mapping of control BYTES to Ethernet channel at 1228.8Mbit/s CPRI line
bit rate and pointer BYTE #Z.194.0=rr111111
4.2.7.7.3. Minimum C&M Channel Support
The use of either HDLC or Ethernet is optional. It is recommended for each REC or RE to support at least
one non-zero C&M channel bit rate on at least one link.
4.2.7.7.4. Passive Link
A passive link does not support any C&M channel. It may be requested by the master port indicating #Z.66.0
= rrrr r000 and #Z.194.0 = rr00 0000 (r = reserved, transmit 0, receiver dont care) in downlink.

CPRI
CPRI Specification V4.2 (2010-09-29) 55
4.2.7.8. Future Protocol Extensions
There are 52 control words of one hyperframe reserved for future interface protocol extensions. Reserved
words are completely filled with reserved bits (reserved bits are marked with r). This means that a
transmitter shall send 0s for bits marked with r, and the receiver shall not interpret bits marked with r.
(transmit: r = 0, receiver: r = dont care).
4.2.7.9. Vendor Specific Data
Depending on the usage of the fast C&M channel up to 192 control words (in subchannels 16 to 63) of one
hyperframe are available for vendor specific data. A minimum of 16 control words (in subchannels 16 to 19)
per hyperframe are reserved for vendor specific data.
4.2.8. Synchronisation and Timing
The RE shall use the incoming bit clock at the slave port where the SAP
S
is assigned as the source for the
radio transmission and any link transmission bit clock. The time information is transferred from the REC to
the RE through the information described in Section 4.2.7.5. The CPRI 10ms frame delimitation is provided
by the K28.5 symbol of the hyperframe number #0.
4.2.8.1. UMTS frame timing
The UMTS radio frame is identical to the CPRI 10ms frame.
In this document the term "UMTS radio frame" is used for the UTRA FDD 10ms frame as well as for the E-
UTRA 10ms frame.
4.2.8.2. WiMAX frame timing
The WiMAX frame timing is defined relative to CPRI 10ms frame timing per AxC or AxC Group. Uplink and
downlink may have different WiMAX frame timing
7
.
The WiMAX frame per AxC Group in a CPRI link is typically aligned with CPRI 10ms frame, especially in the
non-networking case, but may not be aligned with the CPRI 10ms frame and may not be aligned with the
WiMAX frame of other AxC Groups in general, especially in the networking case. The REC informs the RE
about the timing offset between the CPRI frame and the WiMAX frame per AxC Group via the C&M plane
channel. The offset is defined as follows and shown in the Fig. 23A. As the length of a WiMAX frame is the
integer multiple of the CPRI basic frame (e.g. 5ms = 19200 CPRI basic frames), the frame boundary of each
WiMAX frame is identified by this offset and WiMAX frame length in CPRI basic frames.
WiMAX Frame Offset:
The timing difference between the first CPRI basic frame (the basic frame number #0, the hyperframe
number #0 and the BFN number #0) and the first basic frame of the WiMAX Frame assigned to the AxC
Group.
The first basic frame of the WiMAX Frame is always aligned with the first basic frame of an AxC Container
Block. The WiMAX frame duration is an integer multiple of the AxC Container Block duration.


7
This WiMAX frame timing is not the actual WiMAX frame timing of the air interface but is the reference timing between REC and RE in
WiMAX timing domain. This is similar to BFN in UMTS which is not identical to SFN or CFN.

CPRI
CPRI Specification V4.2 (2010-09-29) 56

Figure 23A: WiMAX frame offset within CPRI frame timing

4.2.9. Link Delay Accuracy and Cable Delay Calibration
8

The interface provides the basic mechanism to enable calibrating the cable delay on links and the round trip
delay on multi-hop connections. More specifically, the reference points for delay calibration and the timing
relation between input and output signals at RE are defined. All definitions and requirements in this section
are described for a link between REC and RE. However, it shall also apply for links between two REs if the
master port of the REC is replaced by a master port of a RE.

4.2.9.1. Definition of Reference Points for Cable Delay Calibration
The reference points for cable delay calibration are the input and the output points of the equipment, i.e. the
connectors of REC and RE as shown in Figure 24 and Figure 24A. Figure 24 shows the single-hop
configuration and Figure 24A shows the multi-hop configuration.
Reference points R1-4 correspond to the output point (R1) and the input point (R4) of REC, and the input
point (R2), and the output point (R3) of an RE terminating a particular logical connection between SAP
IQ
. The
antenna is shown as Ra for reference.
REC RE
T12
R1
R2
R4 R3
T34
Toffset
T2a
Ta3
Ra
T14

Figure 24: Definition of reference points for delay calibration (single-hop configuration)
Reference points RB1-4 in the networking RE correspond to the input point (RB2) and the output point (RB3)
of the slave port and the output point (RB1) and the input point (RB4) of the master port.

8
This section describes the single-hop configuration and the multi-hop configurations with networking RE(s) only. This section may be
applied to any other multi-hop configurations including networking REC(s). See section 6.3.8 for further explanation.
CPRI frame timing
sync byte
hyper frame #
BFN # 0
0 1 ... 149
WiMAX frame timing
AxC container
block
WiMAX
Frame
Offset
WiMAX Frame
boundary
0 1 2 3 ...
WiMAX Frame
boundary
WiMAX Frame (T
F
)
0 1 2 3 ...
basic frame # 0 , 1 , 2 , , 255
T
F
/f
S
-1 T
F
/f
S
-1

CPRI
CPRI Specification V4.2 (2010-09-29) 57
REC
Toffset
T12
(1)
R1
R4
T34
(1)
T14
(1)
T12
(2)
T34
(2)
RE
R2
R3
T2a
Ta3
Ra
m
a
s
t
e
r

p
o
r
t
s
l
a
v
e

p
o
r
t
networking
RE
Toffset
(1)
RB2
RB3
RB1
RB4
TBdelay DL
(1)
TBdelay UL
(1)
m
a
s
t
e
r

p
o
r
t
s
l
a
v
e

p
o
r
t
Figure 24A: Definition of reference points for delay calibration (multi-hop configuration)
4.2.9.2. Relation between Downlink and Uplink Frame Timing
Any RE shall use the incoming frame timing at the slave port where SAP
S
is assigned as synchronization
source (RB2 and R2, respectively) as the timing reference for any outgoing signals. The timing specifications
are defined as follows. The single-hop case is explained first using Figure 25, then the multi-hop case is
explained using Figure 25A.
Figure 25 shows the relation between downlink and uplink frame timing for the single-hop configuration.
T12 is the delay of downlink signal from the output point of REC (R1) to the input point of RE (R2).
T34 is the delay of uplink signal from the output point of RE (R3) to the input point of REC (R4).
Toffset is the frame offset between the input signal at R2 and the output signal at R3.
T14 is the frame timing difference between the output signal at R1 and the input signal at R4.

RE shall determine the frame timing of its output signal (uplink) to be the fixed offset (Toffset) relative to the
frame timing of its input signal (downlink). This fixed offset (Toffset) is an arbitrary value, which shall be
greater than or equal to 0 and less than 256 T
C
. In case the system shall fulfil R-21 and R-21A (delay
calibration) then Toffset accuracy shall be better than 8.138ns (=T
C
/32). Different REs may use different
values for Toffset. REC shall know the value of Toffset of each RE in advance (e.g. pre-defined value or RE
informs REC by higher layer message). In addition, the downlink BFN and HFN from REC to RE shall be
given back in uplink from the RE to the REC. In case of an uplink signalled LOS, LOF, RAI or SDI the REC
shall treat the uplink BFN and HFN as invalid.
BFN=0, HFN=0 BFN=0, HFN=1 R1: REC output
BFN=0, HFN=0 BFN=0, HFN=1 R2: RE input
BFN=0, HFN=0 BFN=0, HFN=1 R3: RE output
BFN=0, HFN=0 BFN=0, HFN=1 R4: REC input
T12
Toffset
T34
T14
sync byte

Figure 25: Relation between downlink and uplink frame timing (single-hop configuration)
Figure 25A shows the relation between downlink and uplink frame timing for multi-hop configuration.
The end-to-end delay definitions (T12, T34 and T14) and the frame timing offset (Toffset) for a multi-hop
connection are the same as those of the single-hop configuration.

CPRI
CPRI Specification V4.2 (2010-09-29) 58
The delay of each hop, the frame timing offset and the internal delay in each networking RE are defined as
follows:
M is the number of hops for the multi-hop connection, where M>=2.
T12
(i)
, T34
(i)
and T14
(i)
(1<=i<=M) is the delay of downlink signal, the delay of uplink signal and the frame
timing difference between downlink and uplink of i-th hop respectively.
Toffset
(i)
(1<=i<=M) is the frame offset between the input signal at RB2 and the output signal at RB3 of
the i-th RE. Toffset
(M)
= Toffset.
Tbdelay DL
(i)
(1<=i<=M-1) is the delay of downlink signal between RB2 and RB1 of the i-th networking
RE.
Tbdelay UL
(i)
(1<=i<=M-1) is the delay of uplink signal between RB4 and RB3 of the i-th networking RE.
The timing specifications are as follows:
The same rule is applied for Toffset
(i)
(1<=i<= M) as for Toffset of a single-hop configuration.
Each networking RE shall determine the frame timing of its output signal (downlink) at RB1 to be the fixed
delay (Tbdelay DL
(i)
) relative to the frame timing of its input signal (downlink) at RB2. The frame position
of downlink AxC Container (BFN, HFN and basic frame number) shall be kept unchanged. The position
of AxC Container in a basic frame may be changed.
Each networking RE may change the frame position (BFN, HFN and basic frame number) of uplink AxC
Container carrying a particular IQ sample(s) to minimize the delay between RB4 and RB3. (This is
applicable only when the contents in AxC Containers are not modified, i.e. the bit position of a particular
IQ sample in AxC Container is kept unchanged). The difference of the frame position at RB3 relative to
RB4 transferring the same uplink AxC Container shall be reported to the REC. The unit of the difference
of frame positions is basic frame. In Figure 25A, the AxC Container in the frame position (BFN=0,
HFN=0 and basic frame number=0) at RB4 is transferred in the frame position (BFN=0, HFN=0 and basic
frame number=N
(i)
). In this case the networking RE shall report the value N
(i)
to the REC as the
difference of frame positions of uplink AxC Container.
The end-to-end frame timing difference T14 has the following relation with the 1
st
hop frame timing
difference T14
(1)
:
T14= T14
(1)
+ N x T
C
, where T
C
is the basic frame length and N is calculated as

=
=
1
1
) (
M
i
i
N N .

CPRI
CPRI Specification V4.2 (2010-09-29) 59

Figure 25A: Relation between downlink and uplink frame timing (multi-hop configuration)

4.2.9.3. Definition of Reference Points for Link Delay Accuracy
The reference points for the link delay accuracy and the round trip delay accuracy according to baseline
requirements R-19 and R-20, respectively, are the service access points SAP
S
. The cable delays with their
reference points, as defined in section 4.2.9.1, are excluded from the link delay accuracy requirements. In
case the system shall fulfil R-19 (link delay accuracy) then the accuracy of TbdelayUL
(i)
and TbdelayDL
(i)

which the REC is informed about shall be better than 8.138ns (=T
C
/32).

4.2.10. Link Maintenance of Physical Layer
4.2.10.1. Definition
Four layer 1 alarms are defined
Loss of Signal (LOS)
Loss of Frame (LOF)
Remote Alarm Indication (RAI)
SAP Defect Indication (SDI)

CPRI
CPRI Specification V4.2 (2010-09-29) 60
For each of these alarms a bit is allocated in the CPRI hyperframe to remotely inform the far-end equipment
of the occurrence of the alarm.
On detection of the alarm at near end the inband bit is immediately up to the performance of the device-
set and forwarded on CPRI to the far end. When the alarm is cleared the inband bit is reset.
Notice that to be able to receive and decode such information, the remote equipment must be at least in
state C of start-up (for state definition, see Section 4.5).
Local actions are undertaken at both near and far end when failure is detected.
Failure is:
defined when the alarm persists.
set after time filtering of the alarm.
cleared after time filtering of the alarm.
The timers for near and far end filtering are defined by the application layer.
4.2.10.2. Loss of Signal (LOS)
4.2.10.2.1. Detection
The CPRI definition of LOS is when at least 16 8B/10B violations occur among a whole hyperframe.
For optical mode of CPRI, detection of LOS may also be achieved by detecting light power below a
dedicated threshold. Detection speed shall be within one hyperframe duration.
4.2.10.2.2. Cease
The alarm is cleared when a whole hyperframe is received without code violation.
4.2.10.2.3. Inband Bit
The inband bit that transport this information is #Z.130.0 b3
4.2.10.2.4. Local Action
RE
Upon detecting such a failure, the RE shall go into state B of the start-up sequence (see Section 4.5). In
addition it is HIGHLY recommended that appropriate actions be performed to prevent from emitting on the
radio interface.
REC
On detecting such a failure, the REC shall go into state B of the start-up sequence.
4.2.10.2.5. Remote Action
RE
When detecting such a failure, based on the received information, the RE shall go into state B of the start-up
sequence.
In addition it is HIGHLY recommended that appropriate actions be performed to prevent from emitting on the
radio interface.
REC
When detecting such a failure, based on the received information, the REC shall go into state B of start-up
sequence.

CPRI
CPRI Specification V4.2 (2010-09-29) 61
4.2.10.3. Loss of Frame (LOF)
4.2.10.3.1. Detection
This alarm is detected if the hyperframe alignment cannot be achieved or is lost as shown in Figure 26.
Number of XACQ state and XSYNC state is restricted to acquisition time limitation. Figure 26 shows 2 XACQ
and 3 SYNC states as an example.
XACQ1
XACQ2
XSYNC1
XSYNC2
HFNSYNC
(BYTE=K28.5
& LOS = 0)
(BYTEK28.5 & Y=W=X=0)
(BYTE=K28.5
& Y=W=X=0)
(BYTEK28.5 & Y=W=X=0)
(BYTEK28.5 & Y=W=X=0)
(BYTEK28.5 & Y=W=X=0)
set Y:=W:=X:=0
(BYTE=K28.5
& Y=W=X=0)
(BYTE=K28.5
& Y=W=X=0)
LOS=1
from any state
power-up/reset
(BYTE=K28.5
& Y=W=X=0)
LOF:=1
LOF:=0

Figure 26: Example for LOF and HFNSYNC detection
For receivers with highest available protocol version 2, figure 26A applies instead of figure 26. However, it
may use figure 26 if it receives protocol version 1 from the transmitter.
In the example given in figure 26A 32 bits are used for checking the descrambling sequence.

CPRI
CPRI Specification V4.2 (2010-09-29) 62
XACQ2
XSYNC1
XSYNC2
(BYTE=K28.5
& LOS = 0)
HFNSYNC
(BYTE=K28.5
& Y=W=X=0)
&
(BYTE
(descrambled)
= 50h
& W=X=0 & Y=2..5)
(BYTEK28.5 & Y=W=X=0)
or
(BYTEK28.5 & Y=W=X=0)
or
XACQ1
power-up/reset
From any state
LOS=1
Set Y:= W:=:X:=0
Generate descrambling
sequence
( k [2,..,5] BYTE
(descrambled)
50h
& (W=X=0 & Y=k)
(BYTE=K28.5
& Y=W=X=0)
&
(BYTE
(descrambled)
= 50h
& (W=X=0 & Y=2..5)
(BYTE=K28.5
& Y=W=X=0)
&
(BYTE
(descrambled)
= 50h
& W=X=0 & Y=2..5)
(BYTE=K28.5
& Y=W=X=0)
&
(BYTE
(descrambled)
= 50h
& W=X=0 & Y=2..5)
(BYTEK28.5 & Y=W=X=0)
or
( k [2,..,5] BYTE
(descrambled)
50h
& (W=X=0 & Y=k)
LOF:=1
LOF:=0
( k [2,..,5] BYTE
(descrambled)
50h
& (W=X=0 & Y=k)
(BYTEK28.5 & Y=W=X=0)
or
( k [2,..,5] BYTE
(descrambled)
50h
& (W=X=0 & Y=k)


Figure 26A: Example for LOF and HFNSYNC detection
4.2.10.3.2. Cease
This alarm is cleared if the hyperframe alignment is achieved as shown in Figure 26 and Figure 26A.
4.2.10.3.3. Inband Bit
The inband bit that transports this information is #Z.130.0 b4
4.2.10.3.4. Local Action
RE
When detecting such a failure the RE shall go in state B of start-up sequence.
In addition it is HIGHLY recommended that appropriate actions be performed to prevent emission on the
radio interface.

CPRI
CPRI Specification V4.2 (2010-09-29) 63
REC
When detecting such a failure, based on the received information, the REC shall go in state B of start-up
sequence.
4.2.10.3.5. Remote Action
RE
When detecting such a failure, based on the received information, the RE shall go in state B of start-up
sequence.
In addition it is HIGHLY recommended that appropriate actions be performed to prevent emission on the
radio interface.
REC
When detecting such a failure, based on the received information, the REC shall go in state B of start-up
sequence.
4.2.10.4. Remote Alarm Indication
4.2.10.4.1. Detection
Any errors, including LOS and LOF, that are linked to CPRI transceiver are indicated by the RAI information.
4.2.10.4.2. Cease
When no errors, including LOS and LOF, are linked to the CPRI transceiver, the RAI is cleared.
4.2.10.4.3. Inband Bit
The Remote Alarm Indication bit is used to transport this information: #Z.130.0 b1
4.2.10.4.4. Local Action
RE
Out of scope of CPRI.
REC
Out of scope of CPRI.
4.2.10.4.5. Remote Action
RE
When detecting such a failure, based on the received information, the RE shall go in state B of start-up
sequence.
In addition it is HIGHLY recommended that appropriate actions be performed to prevent from emitting on the
radio interface.
REC
When detecting such a failure, based on the received information, the REC shall go in state B of start-up
sequence.
4.2.10.5. SAP Defect Indication
A link is said to be in alarm when the near end explicitly informs the far end equipment that the link shall not
be used for any of the Service Access Points.
Notice in this case the CPRI link is fully available and decoded by the far end receiver.

CPRI
CPRI Specification V4.2 (2010-09-29) 64
4.2.10.5.1. Detection
The detection procedure is outside the scope of CPRI. This is fully application dependant.
4.2.10.5.2. Cease
The alarm reset procedure is outside the scope of CPRI. This is fully application dependant.
4.2.10.5.3. Inband Bit
The SAP Defect Indication Signal bit is used to transport this information: #Z.130.0 b2
4.2.10.5.4. Local Action
RE
N/A
REC
N/A
4.2.10.5.5. Remote Action
RE
The RE shall not use this link anymore for any of the CPRI Service Access Points: IQ, Sync or C&M. In
addition it is HIGHLY recommended that appropriate actions be performed to prevent from emitting on the
radio interface.
REC
The REC shall not use this link anymore for any of the CPRI Service Access Points: IQ, Sync or C&M.
4.3. Data Link Layer (Layer 2) Specification for Slow C&M Channel
CPRI slow C&M Data Link Layer shall follow the HDLC standard ISO/IEC 13239:2002 (E) [10] using the bit
oriented scheme.
4.3.1. Layer 2 Framing
HDLC data frames and layer 2 procedures shall follow [10]. In addition the CPRI layer 2 for the slow C&M
channel shall fulfil the following additions:
Information Field Length
HDLC information field length in HDLC frames shall support any number of octets.
Bit Transmission Order of the Information Part
HDLC Information field bit transmission order in HDLC frames shall be least significant bit (LSB) first.
Address field
HDLC frames shall use a single octet address field and all 256 combinations shall be available.
Extended address field shall not be used in HDLC data frames.
Frame Format
HDLC data frames shall follow the basic frame format according to ISO/IEC 13239:2002 (E) [10],
chapter 4.1.1
9
.
4.3.2. Media Access Control/Data Mapping
Media Access Control/Data Mapping shall follow chapter 4.2.7.7.1 of this specification.

9
FCS transmission order in HDLC frames shall be most significant bit (MSB) first as defined in the HDLC standard.

CPRI
CPRI Specification V4.2 (2010-09-29) 65
4.3.3. Flow Control
CPRI slow C&M channel flow control shall follow HDLC standard ISO/IEC 13239:2002 (E) [10]. In addition
CPRI layer 2 for the slow C&M channel shall fulfil the following additions:
Flags
HDLC frames shall always start and end with the flag sequence. A single flag must not be used as
both the closing flag for one frame and the opening flag for the next frame.
Inter-frame time fill
Inter-frame time fill between HDLC frames shall be accomplished by contiguous flags.
4.3.4. Control Data Protection/ Retransmission Mechanism
CPRI slow C&M channel data protection shall follow HDLC standard ISO/IEC 13239:2002 (E) [10]. In
addition CPRI layer 2 for the slow C&M channel shall fulfil the following addition:
Frame Check Sequence (FCS)
CPRI slow C&M channel shall support a FCS of length 16 bit as defined in ISO/IEC 13239:2002 (E)
[10].
Retransmission mechanisms shall be accomplished by higher layer signalling.
4.4. Data Link Layer (Layer 2) Specification for Fast C&M Channel
CPRI C&M Fast Data Link Layer shall follow the Ethernet standard as specified in IEEE 802.3-2005 [1].
4.4.1. Layer 2 Framing
Data mapping in layer 2 shall follow section 3. Media access control frame structure of IEEE 802.3-2005
[1].

Figure 27: Layer 2 Framing
Specific CPRI requirements:
Minimum Ethernet frame length and padding:
Due to the specific CPRI framing, no minimum frame length makes any sense for CPRI application. CPRI
does not specify any minimum frame size and does not require frame padding.
The MAC client Data + PAD field length shall range from 1 to 1500 octets.
1-1500 OCTETS

CPRI
CPRI Specification V4.2 (2010-09-29) 66
Extension field:
The extension field shall not be used within CPRI.
4.4.2. Media Access Control/Data Mapping
Layer 2 data mapping in the CPRI frame is performed according to section 4.2.7.7.2 Fast C&M channel of
this specification.
In addition the Ethernet frame shall be controlled and mapped through usage of section 24.2 Physical
Coding SubLayer (PCS) of IEEE 802.3-2005 [1] concerning 100BASE-X.
PCS supports 4 main features that are not all used by CPRI (see Table 13):

Table 13: PCS features used by CPRI
Feature CPRI support
Encoding/Decoding Fully supported by CPRI
Carrier sense detection and collision detection Irrelevant to CPRI
Serialization/deserialization Irrelevant to CPRI
Mapping of transmit, receive, carrier sense and
collision detection
Irrelevant to CPRI

Table 24-4 in 24. Physical Coding SubLayer (PCS) and Physical Medium Attachment (PMA) sublayer, type
100BASE-X of IEEE 802.3-2005 [1] is modified as shown in Figure 28:


CPRI implementation of 100Base-X
PCS
Transmit Receive
Tx_bits[4:0]
Rx_bits[9:0]
MAC interface is not
specified by CPRI
(MII is an option)
CPRI framing as specified in the section
about fast C&M Channel Structure

Figure 28: CPRI implementation of 100BASE-X PCS
The Ethernet MAC frame shall be encoded using the 4B/5B code of 100BASE-X PCS (Physical Coding
Sublayer) as specified in section 24.2 of IEEE 802.3-2005 [1].
The 4B/5B code list shall be according table 24.1 of IEEE 802.3-2005 [1] (see below).

CPRI
CPRI Specification V4.2 (2010-09-29) 67
Table 14: 4B/5B code list (modified Table 24.1 of IEEE 802.3-2005 [1])

MAC Client
Data nibble

The Ethernet frame shall be delineated by the PCS function as shown in Figure 29:

CPRI
CPRI Specification V4.2 (2010-09-29) 68

MAC Client
Ethernet Packet

Figure 29: Physical Layer Stream of 100BASE-X
4.4.3. Flow Control
No flow control is provided for the fast C&M channel.
4.4.4. Control Data Protection/ Retransmission Mechanism
Data protection shall follow section 3.2.8. Frame Check Sequence (FCS) field of IEEE 802.3-2005 [1]. No
retransmission mechanism is specified for Fast C&M channel layer 2.
4.5. Start-up Sequence
This section defines the sequence of actions to be performed by master and slave ports connected by CPRI.
When both the slave port and the master port are in state F or G, the link is in normal operation.
After a reset, any configurable ports of the RE shall be configured as slave ports. All ports of the RE shall
enter state A. All the master ports of the RE shall remain in state A until at least one of the slave ports has
been in state E.

4.5.1. General
The start-up procedure accomplishes two main things:
Synchronization of layer 1: byte alignment and hyper frame alignment
Alignment of capabilities of the master and slave ports: line bit rate, protocol, C&M channel bit rate,
C&M protocol, vendor specific signalling
Since there is no mandatory line bit rate or C&M channel bit rate the master port and slave port shall, during
the start-up procedure, try different configurations until a common match is detected. The common match
does not have to be optimal it shall be considered as just a first contact where capabilities can be
exchanged for a proper configuration to be used in the following communication.
For all states, it is mandatory to always transmit information consistent with the protocol indicated in #Z.2.0
on all control words on sub-channel 1 and sub-channels 3 to 15.
When changing the line bit rate of the transmitted CPRI, the interruption of transmission shall be less than
0.1s. When changing the line bit rate of the received CPRI, the interruption of reception shall be less than
0.1s. The time to reach HFNSYNC for the receiving unit shall be less than 0.2s, given the precondition that
the far-end transmitter is on, they use the same line bit rate and no bit errors occur.

CPRI
CPRI Specification V4.2 (2010-09-29) 69
In the negotiation steps in state C and D the master and slave ports shall sample and evaluate the received
protocol version and C&M channel bit rates at a rate of at least every 0.1 s. The transmitted protocol version
and C&M channel bit rates shall be updated within 0.2 s after the evaluation.

4.5.2. Layer 1 Start-up Timer
The start-up procedure may be endless due to two reasons:
Fault in one of the units
No common layer 1 protocol or C&M channel bit rate or C&M type.
The supervision may be done per state and per cause, but the start-up procedure also specifies a generic
start-up timer which shall be set upon entry of the start-up procedure and shall be cleared when the C&M
channel is established.
If the timer expires the start-up procedure shall be restarted.
The layer 1 start-up timer is activated in transitions 2, 5, 8, 12, 13, 15.
The layer 1 start-up timer is cleared in transitions 6, 9, 10, 11, 14 and in state E when the higher layer C&M
connection is established.
If the layer 1 start-up timer expires, transition 16 shall take place and state B is entered, possibly modifying
the available set of line bit rates and protocols.
The layer 1 start-up timer expiration time is vendor specific.


Standby
L1
synchronization
C/ M plane (L2+)
setup
Operation
C/M plane
disconnected
L1 LOS/LOF/received RAI
REC/RE Shutdown, RE Reset
From any state
Interface and vendor
specific negotiation
Reconfig-
uration
B
1
2
4 5
6 7
10
8
9
Protocol setup
3
11
Passive
link
14
A
C
D
E
F
G
15
L1 start-up timer expired
16
Protocol missmatch
From state D, E, F, G
C&M speed missmatch
From state E, F
12
13
No C&M
C&M proposed

Figure 30: Start-up states and transitions

CPRI
CPRI Specification V4.2 (2010-09-29) 70
4.5.3. State Description
4.5.3.1. State A Standby
Prerequisites:
None
Description:
Waiting to be configured to start up CPRI. No transmission or reception of CPRI. The operator may configure
a suitable start-up configuration (line bit rate, C&M channel characteristics). The master and slave ports may
also have knowledge about a previous successful configuration.
4.5.3.2. State B L1 Synchronization and Rate Negotiation
Prerequisites:
The set of available line bit rate, protocol versions and C&M plane characteristics are known. This may be
the complete set of the unit or a subset based on operator configuration or previous negotiation between the
units (e.g. from state E).
Description:
During this state, the line bit rate of the interface is determined and both master and slave ports reach layer 1
synchronization up to state HFNSYNC.
Interpreted control BYTES:
#Z.0.0, #Z.64.0
#Z.0.2 #Z.0.T/8-1 for ports where protocol version 2 is available (see figure 26A)
Master port actions:
The master port starts to transmit the CPRI at the highest available line bit rate directly when entering the
state, and also start to attempt to receive a CPRI at the same line bit rate. If the master port does not reach
synchronization state HFNSYNC it shall select another line bit rate from CPRI transmission after time T1
from entering the state, given that another line bit rate is available. T1 is 0.9-1.1 s. Each following T1 interval,
a new line bit rate for reception and transmission shall be selected, given that another line bit rate is
available. The line bit rates shall be selected from the available set in a round robin fashion, i.e. first highest,
the second highest, , the slowest, and then restarting from the highest line bit rate.
While in this state, the master port shall set the protocol version in #Z.2.0 to its highest available protocol
version, and the C&M channel bit rates in #Z.66.0 and #Z.194.0 to its highest available C&M channel bit
rates, for the transmitted line bit rate.
Slave port actions:
The slave port shall start attempting to receive CPRI at the highest available line bit rate directly when
entering the state. If the slave port does not reach synchronization state HFNSYNC it shall select another
line bit rate for CPRI reception after T1 from entering the state, given that another line bit rate is available.
T1 is 3.9-4.1s. Each following T1 interval, a new reception line bit rate shall be selected for reception, given
that another line bit rate is available. The line bit rates shall be selected from the available set in a round
robin fashion, i.e. first highest, the second highest, , the slowest, and then restarting from the highest line
bit rate.
When entering this state, the slave port shall turn off its CPRI transmitter, if this state was entered with
transition 10 the slave port may optionally transmit for a maximum of 5 hyperframes to indicate to far-end
equipment the layer 1 link maintenance control BYTE #Z.130.0. When the slave port reaches
synchronization state HFNSYNC, it shall start transmit CPRI on the same line bit rate.
While in this state, the slave port shall set the protocol version in #Z.2.0 according to the rule in state C,
below, or to the highest available protocol version, for the transmitted bit rate. While in this state, the slave
port shall set the C&M channel bit rates in #Z.66.0 and #Z.194.0 according to the rule in state D, or to the
highest available C&M channel bit rate, for the transmitted line bit rate.
Comments:
While in this state, no timer to detect hanging-up is provided by the start-up procedure. Such a hang-up will
occur only in case of HW fault and that is detected by vendor specific means.

CPRI
CPRI Specification V4.2 (2010-09-29) 71
4.5.3.3. State C Protocol Setup
Prerequisites:
Layer 1 is synchronized, i.e., master-to slave and slave-to-master hyper frame structures are aligned.
Description:
During this state, a common protocol version of CPRI is determined.
Interpreted control BYTES:
#Z.0.0, #Z.64.0, #Z.2.0
Master port actions:
The master port shall select its highest available protocol version for the present line bit rate (see table 21)
when entering this state. The protocol version shall be stated in #Z.2.0. When the master port receives a
valid or an updated protocol version from the slave port,
If the currently received protocol version is equal to the current protocol version sent by the master
port, the protocol setup is achieved
If the currently received protocol version differs from the current protocol version sent by the master
port, it shall reselect the protocol version. The new protocol version shall be selected according to
the rule:
New master port protocol version = highest available protocol version which is less or
equal to received slave port protocol version (received in #Z.2.0)
Error case: If no such protocol exists:
New master port protocol version = lowest available protocol version
Note that the reselection may choose the already transmitted protocol version. The new selected protocol
version shall be stated in #Z.2.0. If the currently received protocol version is equal to the new protocol
version sent by the master port, the protocol setup is achieved.
Slave port actions:
The slave port shall decode the received protocol version by looking at #Z.2.0 When the slave port receives
a valid or an updated protocol version from the master port,
If the currently received protocol version is equal to the current protocol version sent by the slave
port, the protocol setup is achieved
If the currently received protocol version differs from the current protocol version sent by the slave
port, the slave port shall reselect the protocol version. The new proposed protocol version shall be
selected according to the rule:
New slave port protocol version = highest available protocol version which is less or equal to
received master port protocol version (received in #Z.2.0)
Error case: If no such protocol exists:
New slave port protocol version = lowest available protocol version
Note that the reselection may choose the already transmitted protocol version. The new selected protocol
version shall be stated in #Z.2.0. If the currently received protocol version is equal to the new protocol
version sent by the slave port, the protocol setup is achieved.
Comments:
If the master port does not receive a new protocol version before the layer 1 start-up timer expires, it can
assume that there are no common protocol versions. Such a detection can be made faster but then the
application must take into account the case where the slave port enters the state after the master port. Layer
1 control bits can start to be interpreted but since they require error protection filtering (majority decision) the
interpretation is not available until the subsequent state D.
4.5.3.4. State D C&M Plane (L2+) Setup
Prerequisites:
Layer 1 is synchronized and the protocol is agreed on.
Description:
During this state, a common C&M channel bit rate is determined.

CPRI
CPRI Specification V4.2 (2010-09-29) 72
Interpreted control BYTES:
All
Master port actions:
The master port shall select its highest available C&M channel bit rate when entering this state: Highest
available HDLC bit rate and highest available Ethernet bit rate. The bit rates shall be stated in #Z.66.0 and
#Z.194.0. When the master port receives a valid or an updated bit rate in either #Z.66.0 or #Z.194.0 from the
slave port,
If at least one of the currently received bit rate is equal to the corresponding bit rate sent by the
master port, the C&M plane setup is achieved
If both currently received bit rates differ from the current bit rates sent by the master port, the master
port shall reselect the C&M channel bit rate in #Z.66.0 and in #Z.194.0. Each new bit rate shall be
selected according to the rule:
New master port bit rate = highest available bit rate which is less or
equal to received slave port bit rate (received in #Z.66.0 or #Z.194.0)
Error case: The resulting bit rate according to the rule is no link, i.e. 0 bit rate:
New master port bit rate = lowest available bit rate
Note that the reselection may choose the already transmitted C&M channel bit rates. The new selected bit
rates shall be stated in #Z.66.0 and #Z.194.0. If at least one of the currently received bit rate is equal to the
corresponding new bit rate sent by master port, the C&M plane setup is achieved.
In this state it is possible for the master port to send #Z.66.0 equal to rrrr r111 if none of the pre-defined
HDLC bit rates are suitable for a specific implementation. This requires that the node is aware in advance of
the characteristics of the HDLC channel when transmitting value rrrr r111 in #Z.66.0.
The master port shall check that #Z.2.0 is equal in both directions. If it is not equal it shall enter state C.
Slave port actions:
The slave port shall decode the received C&M channel bit rates by looking at both #Z.66.0 and #Z.194.0.
When the slave port receives a valid or an updated bit rate in either #Z.66.0 or #Z.194.0 from the master
port,
If at least one of the currently received bit rates is equal to the corresponding bit rate sent by the
slave port, the C&M plane setup is achieved
If both currently received bit rates differ from the current bit rates sent by the slave port the slave port
shall reselect the C&M channel bit rates for each C&M channel, i.e. on both #Z.66.0 and #Z.194.0.
The new proposed C&M channel bit rates shall be selected according to the rule:
New slave port bit rate = highest available bit rate which is less or
equal to received master port bit rate (received in #Z.66.0 or #Z.194.0)
Error case: The resulting bit rate according to the rule is no link, i.e. 0 bit rate:
New slave port bit rate = lowest available bit rate
Note that the reselection may choose the already transmitted C&M channel bit rates. The new selected bit
rates shall be stated in #Z.66.0 and #Z.194.0. If at least one of the currently received bit rates is equal to the
corresponding new bit rate sent by the slave port, the C&M plane setup is achieved.
If the slave port received #Z.66.0 = "rrrr r111" from the master port and if the slave port node is aware in
advance of the characteristics of the HDLC channel, it should send #Z.66.0 equal to "rrrr r111".
The slave port shall check that #Z.2.0 is equal in both directions. If it is not equal it shall enter state C.
Comments:
If the master port does not receive a new C&M channel bit rate proposal before the layer 1 start-up timer
expires, it can assume that there are no common C&M channel bit rates on this line bit rate. Such a
detection can be made faster but then the application must take into account the case where the slave port
enters the state after the master port. The negotiation results in a common C&M channel bit rate on at least
one of the available C&M channels. While in this state, L1 inband protocol is interpreted which may lead to
state G being entered.

CPRI
CPRI Specification V4.2 (2010-09-29) 73
4.5.3.5. State E Interface and Vendor specific Negotiation
Prerequisites:
One C&M channel bit rate is agreed on.
Description:
During this state, application in master and slave ports negotiate the CPRI usage.
Interpreted control BYTES:
All
Master port actions:
If a common bit rate for the Ethernet link was agreed on in state D, it shall be used. Otherwise the HDLC link
shall be used. In this state a negotiation to a HDLC bit rate that is not one of the pre-defined bit rates may
take place. After the negotiation the master port will set #Z.66.0 to rrrr r111 to indicate to the slave port that
a new HDLC bit rate is used, the characteristics of the negotiated HDLC channel is vendor specific. The
connection establishment and higher layer negotiation is outside the scope of the specification. When the
connection is established the layer 1 start-up timer shall be cleared.
The master port shall check that #Z.2.0 is equal in both directions. If it is not equal it shall enter state C. The
master port shall check that at least one of the values #Z.66.0 or #Z.194.0 is equal in both directions. If both
differ, it shall enter state D.
Slave port actions:
If a common bit rate for the Ethernet link was agreed on in state D, it shall be used. Otherwise the HDLC link
shall be used. In this state a negotiation to a HDLC bit rate that is not one of the pre-defined bit rates may
take place. After the negotiation the slave port will set #Z.66.0 to rrrr r111 to indicate to the master port that
a new HDLC bit rate is used, the characteristics of the negotiated HDLC channel is vendor specific. The
connection establishment and higher layer negotiation is outside the scope of the specification. When the
connection is established the layer 1 start-up timer shall be cleared.
The slave port shall check that #Z.2.0 is equal in both directions. If it is not equal it shall enter state C. The
slave port shall check that at least one of the values #Z.66.0 or #Z.194.0 is equal in both directions. If both
differ, it shall enter state D.
Comments:
The master and slave ports exchange information about capabilities and capability limitations resulting in a
preferred configuration of the CPRI, including also the vendor specific parts. The negotiation and the
corresponding C&M messages are not within the scope of the CPRI specification. The result of the
negotiations may require a reconfiguration of the slave or master circuitry. Depending on the degree of
change, the start up procedure may have to restart at state B, C or D, with a new set of characteristics (line
bit rate, protocol, C&M channel bit rate).
4.5.3.6. State F Operation
Prerequisites:
The optimum supported C&M channel is established. The use of the vendor specific area is agreed upon.
Description:
Normal operation.
Interpreted control words:
All
Master port actions:
The master port shall check that #Z.2.0 is equal in both directions. If it is not equal it shall enter state C. The
master port shall check that at least one of the values #Z.66.0 or #Z.194.0 is equal in both directions. If both
differ, it shall enter state D.
Slave port actions:
The slave port shall check that #Z.2.0 is equal in both directions. If it is not equal it shall enter state C. The
slave port shall check that at least one of the values #Z.66.0 or #Z.194.0 is equal in both directions. If both
differ, it shall enter state D.
Comments:
In normal operation, the C&M plane has been established and all further setup of HW, functionality, user

CPRI
CPRI Specification V4.2 (2010-09-29) 74
plane links, IQ format, etc is conducted using procedures outside the scope of the CPRI specification. If the
CPRI is subject to a failure state, B is entered. If a reconfiguration is required state D may be entered.
4.5.3.7. State G Passive Link
Prerequisites:
Layer 1 is synchronized and the protocol is agreed on. The master port does not propose any C&M channel.
Description:
The interface is not carrying the C&M plane
Interpreted control BYTES:
All
Master port actions:
While in this state, the master port shall set the C&M channel bit rates in #Z.66.0 and #Z.194.0 to 0. The
master port shall check that #Z.2.0 is equal in both directions. If not equal it shall enter state C.
Slave port actions:
While in this state, the slave port shall set the C&M channel bit rates in #Z.66.0 and #Z.194.0 to the highest
available bit rate. The slave port shall check #Z.2.0 is equal in both directions. If it is not equal it shall enter
state C. The slave port shall detect any change in the received value #Z.66.0 or #Z.194.0. If at least one
value changes it shall enter state D.
Comments:
This state may be entered due to any of the following reasons:
The interface is used for redundancy and does not carry any information at the moment. Further setup is
done on the active link.
The interface is used to expand the user plane capacity and its I&Q streams are part of the user plane.
Further setup is done on the active link.
As a fallback, the master port may enable the C&M channel by proposing a C&M channel bit rate and the
start-up then enters state D. It is therefore important that the slave port transmits a proper C&M channel bit
rate.
4.5.4. Transition Description
4.5.4.1. Transition 1
Trigger:
The trigger is out of the scope of the CPRI specification. But it is required for the CPRI circuit initiation to be
completed. For the master ports of an RE, this transition is not allowed before one of the slave ports of the
RE has been in state E after reset.
A set of available line bit rates, protocol versions and C&M channel bit rates shall be available. This may be
the equipment full capabilities or a subset determined by the equipment configuration (manual) or knowledge
from previous successful configurations. Such a subset will shorten the time in state B, C and D. Time and
frequency references shall be predictive for the master port.
Actions:
None
4.5.4.2. Transition 2
Trigger:
First time the synchronization state HFNSYNC is entered. Received CPRI line bit rate is equal to transmitted
CPRI line bit rate.
Actions:
The layer 1 start-up timer is set.

CPRI
CPRI Specification V4.2 (2010-09-29) 75
4.5.4.3. Transition 3
Trigger:
Protocol is agreed on. First time transmitted #Z.2.0 is equal to received #Z.2.0.
Actions:
None
4.5.4.4. Transition 4
Trigger:
The C&M channel bit rate is agreed on. First time at least one of the two conditions below is fulfilled:
Received #Z.66.0 is equal to transmitted #Z.66.0, and received #Z.66.0 indicates a valid bit rate.
Received #Z.194.0 is equal to transmitted #Z.194.0, and received #Z.194.0 indicates a valid bit rate.
4.5.4.5. Transition 5
Trigger:
Out of the scope of the CPRI specification. Application has selected a new C&M channel bit rate set and the
C&M channel bit rate is re-setup.
Actions:
The layer 1 start-up timer is set.
4.5.4.6. Transition 6
Trigger:
Out of the scope of the CPRI specification. The capability negotiation is accepted by both master and slave
ports applications and the present CPRI configuration is considered to be the best available choice.
Actions:
The layer 1 start-up timer is cleared.
4.5.4.7. Transition 7
Trigger:
Out of the scope of the CPRI specification. A capability update requiring CPRI capability renegotiation is
performed by the applications.
Actions:
None
4.5.4.8. Transition 8
Trigger:
Out of the scope of the CPRI specification. The C&M plane connection is detected lost by the application due
to fault or reconfiguration.
Actions:
The layer 1 start-up timer is set.
4.5.4.9. Transition 9
Trigger:
Out of the scope of the CPRI specification. The capability negotiation by the application proposes a new
CPRI protocol or line bit rate.
Actions:
The transition carries information about the agreed available set of line bit rates, protocol versions and C&M
channel bit rates. The layer 1 start-up timer is cleared.

CPRI
CPRI Specification V4.2 (2010-09-29) 76
4.5.4.10. Transition 10
Trigger:
First time LOS or LOF or received RAI has been found faulty as defined in 4.2.10.
Actions:
The layer 1 start-up timer is cleared.
4.5.4.11. Transition 11
Trigger:
The slave or master ports are initiated.
Actions:
The layer 1 start-up timer is cleared.
4.5.4.12. Transition 12
Trigger:
First time any of the received C&M channel bit rates in #Z.66.0 or #Z.194.0 is changed while in state E or F.
Actions:
The layer 1 start-up timer is set.
4.5.4.13. Transition 13
Trigger:
First time the received protocol version in #Z.2.0 is changed while in state D, E, F or G.
Actions:
The layer 1 start-up timer is set.
4.5.4.14. Transition 14
Trigger:
First time the master port has set the #Z.66.0 and #Z.194.0 to indicate that no C&M channel is desired on the
interface.
Actions:
The layer 1 start-up timer is cleared.
4.5.4.15. Transition 15
Trigger:
First time the master port proposes C&M channel bit rates in at least one of #Z.66.0 or #Z.194.0.
Actions:
The layer 1 start-up timer is set.
4.5.4.16. Transition 16
Trigger:
When layer 1 start-up timer expires.
Actions:
None


CPRI
CPRI Specification V4.2 (2010-09-29) 77
5. Interoperability
5.1. Forward and Backward Compatibility
5.1.1. Fixing Minimum Control Information Position in CPRI Frame
Structure
For forward and backward compatibility, the minimum control information position shall be fixed in the CPRI
frame in order to find CPRI protocol version correctly. In later versions the position within CPRI hyperframe
of the below listed bits shall not be changed:
Sync and timing (control BYTE: #Z.0.0)
Protocol version (control BYTE: #Z.2.0)
HFN (control BYTE: #Z.64.0)
5.1.2. Reserved Bandwidth within CPRI
Within the CPRI structure some data parts are reserved for future use. These parts may be used in future
releases of the CPRI specification to enhance the capabilities or to allow the introduction of new features in a
backward compatible way.
Two types of reserved blocks need to be distinguished:
Reserved Bits:
Reserved bits are marked with r. This means that a transmitter shall send 0s for bits marked with r, and
the receiver shall not interpret bits marked with r (transmit: r = 0, receiver: r = dont care).
Reserved Control Words:
In the current version of the specification 52 control words (sub channels 3 to 15) of one hyperframe are
reserved for future interface protocol extensions. Reserved words are completely filled with reserved bits
(reserved bits are marked with r).
CPRI reserved data parts shall be used only for protocol enhancements/modifications by the CPRI
specification group.
5.1.3. Version Number
The CPRI specification version is indicated by two digits (version A.B). The following text defines the digits:
The first digit A is incremented to reflect significant changes (modification of the scope, new section)
The second digit B is incremented for all changes of substance, i.e. technical enhancements,
corrections, updates,
5.1.4. Specification Release Version mapping into CPRI Frame
The control BYTE #Z.2.0 indicates the protocol version number, which will be denoted by 1, 2, 3, The
protocol version number will be incremented only when a new specification release version includes changes
that lead to incompatibility with previous specification release versions. The simple sequence and the well-
defined rule for non-compatibility between different specification release versions allow a simple, efficient
and fast start-up procedure. The following table provides the mapping between specification release version
and protocol version number.

CPRI
CPRI Specification V4.2 (2010-09-29) 78

Table 15: Specification release version and protocol version numbering
Specification release version Compatible with the following
previous specification release
versions
Highest available protocol
version number (Z.2.0 control
BYTE)
1.0 - 1
1.1 1.0 * 1
1.2 1.0 *, 1.1 1
1.3 1.0 *, 1.1, 1.2 1
2.0 1.0 *, 1.1, 1.2, 1.3 1
2.1 1.0 *, 1.1, 1.2, 1.3, 1.4, 2.0 1
3.0 1.0 *, 1.1, 1.2, 1.3, 1.4, 2.0, 2.1 1
4.0 1.0 *, 1.1, 1.2, 1.3, 1.4, 2.0, 2.1,
3.0
1
4.1 1.0 *, 1.1, 1.2, 1.3, 1.4, 2.0, 2.1,
3.0, 4.0
1: scrambling not supported
2: scrambling supported
4.2 1.0 *, 1.1, 1.2, 1.3, 1.4, 2.0, 2.1,
3.0, 4.0, 4.1
1: scrambling not supported
2: scrambling supported

* The compatibility between V1.0 and the other specification release versions requires the V1.0 receiver to
tolerate the /I1/ sequence as specified in section 4.2.7.5.
This table shall be updated when new specification release versions become available.
5.2. Compliance
A CPRI compliant interface application fulfils all following requirements:
Establishes and maintains a connection between RE and REC by means of mandatory and optional
parts of the CPRI specification.
Establishes and maintains a connection between RE and REC by means of supporting all mandatory
parts of CPRI specification.
Establishes and maintains a connection between RE and REC by means of selecting at least one
option out of every option list in the CPRI specification.
Does not add any additional options in an option list.
Does not add additional option lists.
Does not produce errors when passing data between SAPs in RE and REC.

It is not required that all the CPRI compatible modules shall meet the full set of requirements defined in the
section 3. The performances of the module can be restricted to a subset of the requirement when some
application is not requiring the full performance of the CPRI specification.
For each CPRI compatible module, the vendor shall explicitly give the compliance list for each item of the
section 3 that are impacted by the module design even if the full specification requirement is not met.

CPRI
CPRI Specification V4.2 (2010-09-29) 79
6. Annex
6.1. Delay Calibration Example (Informative)
This section provides an example for the delay calibration procedure that has been described in Section
4.2.9. The single-hop case is explained first and then the multi-hop case is explained.
In the case of a single-hop configuration the delay between REC and RE (T12 and T34) can be estimated as
follows.
Step 1) Measure T14, the frame timing difference between the output signal at R1 and the input signal at
R4. Assume <T14> is the measured value of T14.
Step 2) Estimate the round trip delay between REC and RE <T12+T34> by subtracting the known value
Toffset from <T14>. <T12+T34> = <T14> - Toffset
Step 3) If the downlink delay (T12) and the uplink delay (T34) are assumed to be the same, the one way
delay can be estimated from the round-trip delay by halving it.
<T12> = <T34> = <T12+T34> / 2 = (<T14> - Toffset) / 2

As these two reference points R1 and R4 are in the same equipment, REC, it is feasible to measure the T14
accurate enough to fulfil the requirement (R-21) in Section 3.
Of course it may be difficult to measure the timing at R1 and R4 directly because the signals at these points
are optical or electrical high speed signals, but it is feasible to measure the timing difference somewhere in
REC (e.g. before and after the SERDES) and to compensate the internal timing difference between
measurement points and R1/R4.
As it is feasible enough to assume that the REC knows the overall downlink delay (T2a) and the uplink delay
(Ta3) in the RE, the REC is able to estimate the overall delay including the delay between REC and RE by
adding <T12> and <T34>. In case of TDD mode, the computation may require further knowledge of the
actual WiMAX frame configuration.
Where,
T2a is the delay from the UMTS frame boundary (UTRA-FDD/E-UTRA) or the WiMAX frame boundary
(WiMAX) of the downlink signal at R2 to the transmit timing at the RE antenna (Ra) of the first IQ sample
carried in the corresponding UMTS frame (UTRA-FDD/E-UTRA) or the corresponding WiMAX frame
(WiMAX).
Ta3 is the delay from the received signal at the RE antenna (Ra) to the UMTS frame boundary (UTRA-
FDD/E-UTRA) or the WiMAX frame boundary (WiMAX) at R3. The I/Q sample of the corresponding
received signal, which is carried as the first I/Q sample in the UMTS frame (UTRA-FDD/E-UTRA) or in the
WiMAX frame (WiMAX), is used to measure the delay.
In the case of WiMAX, the delay may vary depending on the IQ mapping method and the position of the IQ
sample in a WiMAX frame.
Therefore, the WiMAX frame boundary as defined in section 4.2.8 and the IQ sample, which is carried as the
first sample in a WiMAX frame, are selected to define the delay.
In case of WiMAX TDD/E-UTRA TDD, the first IQ sample in a frame may not have valid content (if
transmitter or receiver is inactive). In this case the equivalent delay is measured using any other valid IQ
sample and the fixed timing relation to the frame.


CPRI
CPRI Specification V4.2 (2010-09-29) 80

BFN=0, HFN=0 R2: RE input
Ra: RE antenna
(Tx signal)
Ra: RE antenna
(Rx signal)
BFN=0, HFN=0 BFN=0, HFN=1 R3: RE output
T2a
Ta3
sync byte
BFN=0, HFN=1
the first I/Q sample in the first basic frame
the first I/Q sample in the first basic frame
sync byte

Figure 31: Definition of RE internal delay (UTRA-FDD and E-UTRA)

R2: RE input
Ra: RE antenna
(Tx signal)
Ra: RE antenna
(Rx signal)
R3: RE output
T2a
Ta3
sync byte
the first I/Q sample in a WiMAX frame
sync byte
the first I/Q sample in a WiMAX frame
control word
WiMAX Frame Offset
(downlink)
WiMAX frame boundary
CPRI 10ms frame boundary
AxC Container Block
WiMAX frame (T
F
)
WiMAX frame boundary CPRI 10ms frame boundary
control word
AxC Container Block
WiMAX Frame Offset (uplink)

Figure 31A: Definition of RE internal delay (WiMAX FDD only)

CPRI
CPRI Specification V4.2 (2010-09-29) 81
R2: RE input
Ra: RE antenna
(Tx signal)
Ra: RE antenna
(Rx signal)
R3: RE output
T2a
Ta3
sync byte
I/Q sample used to measure the delay
sync byte
the first I/Q sample in a WiMAX frame
control word
WiMAX Frame Offset
(downlink)
WiMAX frame boundary
CPRI 10ms frame boundary
AxC Container Block
WiMAX frame (T
F
)
CPRI 10ms frame boundary
control word
WiMAX Frame Offset (uplink)
I/Q sample used to measure the delay
WiMAX frame (T
F
)
WiMAX frame boundary
measured
the first I/Q sample in a WiMAX TDD frame (may not have valid content)

Figure 31B: Definition of RE internal delay (WiMAX TDD only)

In case of a multi-hop configuration
10
the round-trip delay between REC and RE (T12+T34) can be
estimated as follows.
Step 1) Measure T14
(1)
, the frame timing difference between the output signal at R1 and the input signal at
R4. Assume <T14
(1)
> is the measured value of T14
(1)
.
Step 2) Estimate the end-to-end frame timing difference T14 by taking into account the difference of frame
positions of uplink IQ samples N. <T14> = <T14
(1)
> + N x T
C
,
where T
C
is the basic frame length = chip period and N is the sum of all N
(i)
reported by i-th
networking RE (1<=i<=M-1), i.e.

=
=
1
1
) (
M
i
i
N N , M is the number of hops.
Step 3) Estimate the round trip delay between REC and RE <T12+T34> by subtracting the known value
Toffset from <T14>. <T12+T34> = <T14> - Toffset
As the difference of frame positions of uplink IQ samples N is the definite value (no accumulation of
measurement error), the accuracy of round-trip delay does not depend on the number of hops.
However, the estimate of the one-way delay is not as simple as in the single-hop case. Dividing <T12+T34>
by 2 may not introduce the one way delay <T12> and/or <T34> because the assumption <T12> = <T34> is
no longer feasible as the internal delays in networking REs, TBdelayDL
(i)
and TBdelayUL
(i)
, included in
<T12> and <T34> may not be the same for uplink and downlink.


= =
+ =
1
1
) (
) (
1
12 12
M
i
i
i
M
i
TBdelayDL T T


= =
+ =
1
1
) (
) (
1
34 34
M
i
i
i
M
i
TBdelayUL T T
TBdelay DL
(i)
does not depend on the link delay so it is a known value for the networking RE.
TBdelay UL
(i)
depends on the link delay so it has to be measured in the field.
There may be several methods to estimate the one-way delay T12 and/or T34, following is one example to
estimate the T12 and T34.

10
This section describes the multi-hop configuration with networking REs only as an example. The same method may be applied to
any other multi-hop configurations including networking REC(s) if the behaviour described in section 4.2.9 is fulfilled.

CPRI
CPRI Specification V4.2 (2010-09-29) 82
Step 4) Each networking RE needs to report the internal delays TBdelayDL
(i)
and TBdelayUL
(i)
to the
REC.
Step 5) The REC needs to estimate the one-way delay T12 and T34 by using <T12+T34> estimated in step
3 and the values {TBdelayDL
(i)
} and {TBdelayUL
(i)
} (1<=i<=M-1) reported by networking REs as
follows:

=
+ > + < >= <
1
1
) ( ) (
2 / )} ( 34 12 { 12
M
i
i i
TBdelayUL TBdelayDL T T T
and

=
> + < >= <
1
1
) ( ) (
2 / )} ( 34 12 { 34
M
i
i i
TBdelayUL TBdelayDL T T T

6.2. Electrical Physical Layer Specification (Informative)
This section and all the following subsections are informative only.

Four electrical variants are recommended for CPRI usage denoted HV (high voltage), LV (low voltage), LV-II
(low voltage II) and LV-III (low voltage III) in Figure 32. The HV variant is guided by 1000Base-CX electrical
interface specified in Clause 39 of IEEE 802.3-2005 [1], but with 100 impedance and adapted to CPRI line
bit rates. The LV variant is guided by the XAUI electrical interface specified in Clause 47 of IEEE 802.3-2005
[1], but adapted to CPRI line bit rates. The LV-II variant is guided by Clause 7 of OIF-CEI02.0 [17], but
adapted to CPRI line bit rates, and with BER requirement of 10
-12
. The LV-III variant is guided by 10GBase-
KR, defined in IEEE 802.3 [22] clause 72.7 and clause 72.8, but adapted to CPRI line bit rates.

The intention is to be able to reuse electrical designs from 1000BASE-CX, XAUI, OIF-CEI or 10GBase-KR
respectively.

All unit intervals are specified with a tolerance of +/- 100 ppm. The worst-case frequency difference between
any transmit and receive clock will be 200 ppm. Note that this requirement is only aiming at achieving a data
BER of 10
-12
through the CPRI link. The CPRI clock tolerance is driven by 3GPP requirements (see 3GPP TS
25.104 [8]).


6.2.1. Overlapping Rate and Technologies
Four different technologies may be used for CPRI with an overlap with respect to CPRI line bit rate ranges.



Figure 32: HV, LV, LV-II and LV-III electrical layer 1 usage

Nothing prevents inter-operating the four electrical variants after bi-lateral tests. Neither does anything
prevent developing a circuit supporting all variants.


CPRI
CPRI Specification V4.2 (2010-09-29) 83
6.2.2. Signal Definition
The CPRI link uses differential signalling. Figure 33 defines terms used in the description and specification of
the CPRI differential signal pair.

Caution should be taken that some standards and IC data sheet define electrical characteristic with Vdiffpp
value, which is twice Vdiff.
The single ended voltage swing is what is measured on one line of the paired differential signal.

T+
T-
Vhigh
Vlow
+Vdiff=(Vhigh- Vlow)
-Vdiff=(Vlow - Vhigh)
Vdiffpp = 2x Vdiff
Single ended value Differential value
1 1 0
1
0
(T+) (T-)


Figure 33: Definition of differential signals of a transmitter or receiver

6.2.3. Eye Diagram and Jitter
Jitter values and differential voltage levels at both Transmitter and Receiver are specified according to the
reference eye diagram in Figure 34.


X1 1-X1 1-X2 X2
0 V
+min Vdiff
-min Vdiff
+maxVdiff
-max Vdiff
0 1

Figure 34: Definition of eye diagram mask

In addition, deterministic and total jitter budget values are specified.



CPRI
CPRI Specification V4.2 (2010-09-29) 84
6.2.4. Reference Test Points
Four reference test points are specified:



Serdes
Passive/active
elements
Connector
TP1
TP4
TP3
TP2
Transmission
network

Serdes
Transmission
network
Passive/active
elements
Connector

Figure 35: Reference test points

TX and RX requirements are specified at TP1 and TP4 respectively for the Low voltage electrical interface
guided by XAUI. The characteristics of the channel between TP1 and TP4 are not included in the CPRI
specification.
TX and RX requirement are specified at TP2 and TP3 respectively for the High voltage electrical interface
guided by 1000Base-CX. The characteristics of the channel between TP2 and TP3 are not included in the
CPRI specification.
TX and RX requirements are specified at TP1 and TP4 respectively for the LV-II electrical interface guided
by CEI-6G-LR. The characteristics of channel between TP1 and TP4 which can be designed guided by
section 7.3.7 Channel Compliance of OIF-CEI02.0 [17], are not included in the CPRI specification.
TX and RX requirements are specified at TP1 and TP4 respectively for the LV-III electrical interface guided
by 10GBase-KR. The characteristics of channel between TP1 and TP4 which can be designed guided by
IEEE 802.3 [22] section 72.8 Interconnect characteristics, are not included in the CPRI specification.
6.2.5. Cable and Connector
Neither cables, nor PCBs, nor connectors are specified for the CPRI.

6.2.6. Impedance
Four options are specified:
Low Voltage variant: Guided by IEEE 802.3-2005 [1], clause 47. The differential impedance of the
channel is 100 .
High Voltage variant: Guided by IEEE 802.3-2005 [1], clause 39, except that 150 differential
impedance is replaced by 100 .
Low Voltage II variant: Guided by OIF-CEI-02.0, clause 7. The differential impedance of the channel
is 100 .
Low Voltage III variant: Guided by IEEE 802.3 [22], clause 72.7 and Clause 72.8. The differential
impedance of the channel is 100 .
6.2.7. AC Coupling
Four options are specified:

CPRI
CPRI Specification V4.2 (2010-09-29) 85
Low Voltage variant: Guided by IEEE 802.3-2005 [1], clause 47. The link is AC coupled at the
receiver side.
High Voltage variant: Guided by IEEE 802.3-2005 [1], clause 39. The link is AC coupled at the
receiver side and optionally AC coupled at the transmitter side.
Low Voltage II variant: Guided by OIF-CEI-02.0, clause 7. The link is AC coupled at the receiver side
and optionally AC coupled at the transmitter side.
Low Voltage III variant: Guided by IEEE 802.3 [22], clause 72.7. The link is AC coupled at the
receiver side and optionally AC coupled at the transmitter side.
6.2.8. TX Performances
6.2.8.1. LV TX
The serial transmitters electrical and timing parameters for E.6.LV, E.12.LV ,E.24.LV and E.30LV are stated
in this section. All given TX parameters are referred to TP1. The TX parameters are guided by XAUI
electrical interface (IEEE 802.3-2005 [1], clause 47).


0.175 0.825 0.61 0.39
0 V
+ 400 mV
- 400 mV
+ 800 mV
- 800 mV
0 1

Figure 36: E.6.LV, E.12.LV, E.24.LV, E.30.LV transmitter output mask
Table 16: E.6.LV, E.12.LV, E.24.LV and E.30.LV transmitter AC timing specification
Range
Characteristic Symbol
Min Max
Unit Notes
Output Voltage Vo -0.40 2.30 Volts Voltage relative to common of
either signal comprising a
differential pair
Differential Output Voltage V
DIFFPP
800 1600 mV,p-p
Deterministic Jitter J
D
0.17 UI
Total Jitter J
T
0.35 UI
Unit Interval E.6.LV UI 1/614.4 1/614.4 s +/- 100 ppm
Unit Interval E.12.LV UI 1/1228.8 1/1228.8 s +/- 100 ppm
Unit Interval E.24.LV UI 1/2457.6 1/2457.6 s +/- 100 ppm
Unit Interval E.30.LV UI 1/3072.0 1/3072.0 s +/- 100 ppm

The differential return loss, S11, of the transmitter in each case shall be better than
-10 dB for [CPRI line bit rate/10] < f < 625 MHz, and

CPRI
CPRI Specification V4.2 (2010-09-29) 86
-10 dB + 10xlog(f / (625

MHz)) dB for 625 MHz <= f <= [CPRI line bit rate]
The reference impedance for the differential return loss measurement is 100 resistive. Differential return
loss includes contribution from SERDES on-chip circuitry, chip packaging and any off-chip components
related to the driver. The output impedance requirement applies to all valid output levels.
It is recommended that the 20%-80% rise/fall time of the CPRI-LV Serial transmitter, as measured at the
transmitter output, in each case have a minimum value of 60 ps.
It is recommended that the timing skew at the output of a CPRI-LV Serial transmitter between the two signals
that comprise a differential pair does not exceed 15 ps.
6.2.8.2. HV TX
The TX electrical and timing parameters for E.6.HV and E.12.HV are stated in this section. All given TX
parameters are referred to TP2. The TX parameters are guided by 1000Base-CX (IEEE 802.3-2005 [1],
clause 39, PMD to PMI interface).


0.14 0.86 0.66 0.34
0 V
+ 550 mV
- 550 mV
+1000 mV
- 1000 mV
0 1

Figure 37: E.6.HV and E.12.HV transmitter mask

Table 17: E.6.HV and E.12.HV transmitter AC timing specification
Range
Characteristic Symbol
Min Max
Unit Notes
Differential Output Voltage V
DIFFPP
1100 2000 mV,p-p
Rise / Fall time (20% to 80 %) T
RF
85 327 ps
Deterministic Jitter J
D
0.14 UI
Total Jitter J
T
0.279 UI
Output skew S
O
25 ps
Unit Interval E.6.HV UI 1/614.4 1/614.4 s +/- 100 ppm
Unit Interval E.12.HV UI 1/1228.8 1/1228.8 s +/- 100 ppm

The differential return loss, S11, of the transmitter in each case shall be better than
-15 dB for [CPRI line bit rate/10] < f < 625 MHz, and
-15 dB + 10xlog(f / (625

MHz)) dB for 625 MHz <= f <= [CPRI line bit rate]
The reference impedance for the differential return loss measurement is 100 resistive. Differential return
loss includes contribution from SERDES on-chip circuitry, chip packaging and any off-chip components or

CPRI
CPRI Specification V4.2 (2010-09-29) 87
transmission lines related to the driver transmission network. The output impedance requirement applies to
all valid output levels.
6.2.8.3. LV-II TX
The serial transmitters electrical and timing parameters for LV-II are stated in this section. All given TX
parameters are referred to TP1. The TX parameters are guided by CEI-6G-LR electrical interface (OIF-CEI-
02.0 [17], clause 7).

0.15 0.85 0.6 0.4
0 V
+ 400 mV
- 400 mV
+ 600 mV
- 600 mV
0 1

Figure 37A: LV-II transmitter output mask
Table 18A: LV-II transmitter AC timing specification
Range
Characteristic Symbol
Min Max
Unit Notes
Output Voltage Vo 0.1 1.70 Volts Voltage relative to common of
either signal comprising a
differential pair
Differential Output Voltage VDIFFPP 800 1200 mV,p-p
Uncorrelated Bounded High
Probability Jitter
T_UBHP
J
0.15 UI
Duty Cycle Distortion T_DCD 0.05 UI DCD is only required for line
rate 4.9152Gbps
Total Jitter (Peak-to-Peak) J
T
0.30 UI @ 10
-12
BER
Unit Interval E.6.LV-II UI 1/614.4 1/614.4 s +/- 100 ppm
Unit Interval E.12.LV-II UI 1/1228.8 1/1228.8 s +/- 100 ppm
Unit Interval E.24.LV-II UI 1/2457.6 1/2457.6 s +/- 100 ppm
Unit Interval E.30.LV-II UI 1/3072 1/3072 s +/- 100 ppm
Unit Interval E.48.LV-II UI 1/4915.2 1/4915.2 s +/- 100 ppm
Unit Interval E.60.LV-II UI 1/6144.0 1/6144.0 s +/- 100 ppm
.

The DC differential resistance shall be between 80 and 120.
The differential return loss, S11, of the transmitter in each case shall be better than
-8 dB for 100MHz < f < 0.75* [CPRI line bit rate], and

CPRI
CPRI Specification V4.2 (2010-09-29) 88
-8dB + 16.6*log(f / (0.75* [CPRI line bit rate]) ) dB for 0.75* [CPRI line bit rate] <= f <= [CPRI line bit rate]
The reference impedance for the differential return loss measurement is 100 resistive. Differential return
loss includes contribution from SERDES on-chip circuitry, chip packaging and any off-chip components
related to the driver. The output impedance requirement applies to all valid output levels.
The Common Mode Return Loss of the transmitter in each case shall be better than
-6 dB for 100MHz < f < 0.75* [CPRI line bit rate]
The reference impedance for the common mode return loss is 25.
The recommended minimum differential rise and fall time is 30ps as measured between the 20% and 80% of
the maximum measured levels. The maximum differential rise and fall times are defined by the Tx eye
diagram. Shorter rise and falls may result in excessive high frequency components and increase EMI and
cross talk.
It is recommended that the timing skew at the output of a serial transmitter between the two signals that
comprise a differential pair does not exceed 15 ps.

6.2.8.4. LV-III TX
The serial transmitters electrical and timing parameters for LV-III are stated in this section. All given TX
parameters are referred to TP1. The TX parameters are guided by 10GBase-KR electrical interface (IEEE
802.3 [22], clause 72.7.1).

0.15 0.85 0.6 0.4
0 V
+ 400 mV
- 400 mV
+ 600 mV
- 600 mV
0 1

Figure 37B: LV-III transmitter output mask

CPRI
CPRI Specification V4.2 (2010-09-29) 89
Table 18B: LV-III transmitter AC timing specification
Range
Characteristic Symbol
Min Max
Unit Notes
Common-mode voltage limits Vo 0 1.90 Volts
Differential Output Voltage VDIFFPP 800 1200 mV,p-p
Deterministic Jitter T_DJ 0.15 UI
0.005 UI
DCD 0.05 UI (4.9152 rate < 9.8304
Gbps)
Duty Cycle Distortion
T_DCD

0.035 UI DCD 0.035 UI (9.8304 Gbps rate)
Random Jitter T_RJ 0.15 UI @ 10
-12
BER
Unit Interval E.24.LV-III UI 1/2457.6 1/2457.6 s +/- 100 ppm
Unit Interval E.30.LV-III UI 1/3072 1/3072 s +/- 100 ppm
Unit Interval E.48.LV-III UI 1/4915.2 1/4915.2 s +/- 100 ppm
Unit Interval E.60.LV-III UI 1/6144.0 1/6144.0 s +/- 100 ppm
Unit Interval E.96.LV-III UI 1/9830.4 1/9830.4 s +/- 100 ppm

The differential return loss, S11, of the transmitter in each case shall be better than
-9 dB for 50MHz <= f <2500MHz, and
-9dB + 12*log(f / 2500MHz) dB for 2500MHz <= f <= 7500MHz
The reference impedance for the differential return loss measurement is 100 resistive. Differential return
loss includes contribution from SERDES on-chip circuitry, chip packaging and any off-chip components
related to the driver. The output impedance requirement applies to all valid output levels.
The Common Mode Return Loss of the transmitter in each case shall be better than
-6 dB for 50MHz <= f <2500MHz
-6dB + 12*log(f / 2500MHz) dB for 2500MHz <= f <= 7500MHz
The reference impedance for the common mode return loss is 25.
The rising and falling edge transition times shall be between 24 ps and 47 ps as measured at the 20% and
80% levels. Shorter rise and falls may result in excessive high frequency components and increase EMI and
cross talk.
It is recommended that the timing skew at the output of a serial transmitter between the two signals that
comprise a differential pair does not exceed 9 ps.
6.2.8.5. Pre-emphasis and TX-Compliance
Pre-emphasis is allowed by CPRI to overcome data dependent jitter issue. Neither specific pre-emphasis
value nor other equalization technique is specified within CPRI.
The output eye pattern of a CPRI transmitter that implements pre-emphasis (to equalize the link and reduce
inter-symbol interference) need only comply with the Transmitter Output Compliance Mask when pre-
emphasis is disabled or minimized.
For LV and HV variants, Pre-emphasis techniques are to be tested on a bilateral end-to-end basis in
between CPRI Nodes.
For LV-II variant, the Pre-emphasis compliance testing is guided by section 2.4.3 Transmitter Interoperability
of OIF-CEI02.0 [17]. It shall be verified that the measured eye is equal or better than the calculated eye for
the given measurement probability Q (for 10
-12
BER, Q is 7.035).

CPRI
CPRI Specification V4.2 (2010-09-29) 90
For LV-III variant, the Pre-emphasis compliance testing is guided by IEEE 802.3 [22] section 72.7.1.11
Transmitter output waveform requirements of 10GBase-KR.
6.2.9. Receiver Performances
6.2.9.1. LV RX
The serial receiver electrical and timing parameters for E.6.LV, E.12.LV, E.24.LV and E.30.LV are stated in
this section. All given RX parameters are referred to TP4. The RX parameters are guided by XAUI (IEEE
802.3-2005 [1], section 47).


0.275 0.725 0.6 0.4
0 V
+ 100 mV
- 100 mV
+ 800 mV
- 800 mV
0 1

Figure 38: E.6.LV, E.12.LV, E.24.LV and E.30.LV receiver mask
Table 19: E.6.LV, E.12.LV, E.24.LV, and E.30.LV receiver AC timing specification
Range
Characteristic Symbol
Min Max
Unit Notes
Differential Input Voltage V
IN
200 1600 mV,p-p Measured at receiver
Deterministic Jitter J
D
0.37 UI Measured at receiver
Combined Deterministic and
Random Jitter
J
DR
0.55 UI Measured at receiver
Total Jitter J
T
0.65
1
UI Measured at receiver
Bit Error Rate BER 10
-12

Unit Interval E.6.LV UI 1/614.4 1/614.4 s +/- 100 ppm
Unit Interval E.12.LV UI 1/1228.8 1/1228.8 s +/- 100 ppm
Unit Interval E.24.LV UI 1/2457.6 1/2457.6 s +/- 100 ppm
Unit Interval E.30.LV UI 1/3072.0 1/3072.0 s +/- 100 ppm
Note:
1. Total random jitter is composed of deterministic jitter, random jitter and single frequency sinusoidal jitter.
The sinusoidal jitters amplitude and frequency is defined in agreement with XAUI specification IEEE
802.3-2005 [1], clause 47.

Input impedance is defined as 100 and is tested by return loss measurement.
Receiver input impedance shall result in a differential return loss better that 10 dB and a common mode
return loss better than 6 dB from [CPRI line bit rate/10] to [CPRI line bit rate] frequency. This includes

CPRI
CPRI Specification V4.2 (2010-09-29) 91
contributions from on chip circuitry, the chip package and any off-chip components related to the receiver.
AC coupling components are included in this requirement. The reference impedance for return loss
measurements is 100 resistive for differential return loss and 25 resistive for common mode.

6.2.9.2. HV RX
The RX electrical and timing parameters for E.6.HV and E.12.HV are stated in this section. All given RX
parameters are referred to TP3. The RX parameters are guided by 1000Base-CX (IEEE 802.3-2005 [1],
clause 39, PMD to PMI interface).


0.33 0.67 0.5 0.5
0 V
+ 200 mV
- 200 mV
+ 1000 mV
- 1000 mV
0 1

Figure 39: E.6.HV and E.12.HV receiver mask

Table 20: E.6.HV and E.12.HV receiver AC timing specification
Range
Characteristic Symbol
Min Max
Unit Notes
Differential Input Voltage V
IN
400 2000 mV,p-p
Deterministic Jitter J
D
0.40 UI
Total Jitter J
T
0.66 UI
Differential input skew S
I
175 ps
Bit Error Rate BER 10
-12

Unit Interval E.6.HV UI 1/614.4 1/614.4 s +/- 100 ppm
Unit Interval E.12.HV UI 1/1228.8 1/1228.8 s +/- 100 ppm

Input impedance is defined as 100 and is tested by return loss measurement.
Receiver input impedance shall result in a differential return loss better that 15 dB and a common mode
return loss better than 6 dB from [CPRI line bit rate/10] to [CPRI line bit rate] frequency. This includes
contributions from SERDES on chip circuitry, the chip package and any off-chip components or transmission
lines related to the receiver transmission network. AC coupling components are included in this requirement.
The reference impedance for return loss measurements is 100 resistive for differential return loss and 25
resistive for common mode.


CPRI
CPRI Specification V4.2 (2010-09-29) 92
6.2.9.3. LV-II RX
The serial receiver electrical and timing parameters for LV-II are stated in this section. All given RX
parameters are referred to TP4. The RX parameters are guided by CEI-6G-LR electrical interface (OIF-CEI-
02.0, clause 7).

Table 19A: E.48 and E.60 receiver characteristic
Range
Characteristic Symbol
Min Max
Unit Notes
Differential Input Voltage VIN 1200 mV,p-p Measured at receiver
Differential Resistance R_Rdin 80 120
Differential Input Return Loss
100MHz to 0.75*R_Baud)
-8 dB
Differential Input Return Loss
(0.75*R_Baud to R_Baud))
R_SDD11
16.6 dB/dec
Common Mode Input Return
Loss
(100MHz to 0.75 *R_Baud)
R_SCC11 -6 dB
Unit Interval E.6.LV-II UI 1/614.4 1/614.4 s +/- 100 ppm
Unit Interval E.12.LV-II UI 1/1228.8 1/1228.8 s +/- 100 ppm
Unit Interval E.24.LV-II UI 1/2457.6 1/2457.6 s +/- 100 ppm
Unit Interval E.30.LV-II UI 1/3072 1/3072 s +/- 100 ppm
Unit Interval E.48.LV.LV-II UI 1/4915.2 1/4915.2 s +/- 100 ppm
Unit Interval E.60.LV.LV-II UI 1/6144.0 1/6144.0 s +/- 100 ppm

The differential return loss of the receiver shall be better than
-8 dB for 100MHz < f < 0.75* [CPRI line bit rate], and
-8dB + 16.6*log(f / (0.75* [CPRI line bit rate]) ) dB for 0.75* [CPRI line bit rate] <= f <= [CPRI line bit rate]
The reference impedance for the differential return loss measurement is 100 resistive.
The Common Mode Return Loss of the transmitter in each case shall be better than
-6 dB for 100MHz < f < 0.75* [CPRI line bit rate]
The reference impedance for the common mode return loss is 25.
Jitter tolerance is defined in section 6.2.9.5 Equalization and RX Compliance.

6.2.9.4. LV-III RX
The serial receiver electrical and timing parameters for LV-III are stated in this section. All given RX
parameters are referred to TP4. The RX parameters are guided by 10GBase-KR electrical interface (IEEE
802.3 [22], clause 72.7.2).

CPRI
CPRI Specification V4.2 (2010-09-29) 93
Table 19B: E.96 receiver characteristic
Range
Characteristic Symbol
Min Max
Unit Notes
Differential Input Voltage VIN 1200 mV,p-p Measured at receiver
Bit Error Ratio BER 1.0E-12
Unit Interval E.24.LV-III UI 1/2457.6 1/2457.6 s +/- 100 ppm
Unit Interval E.30.LV-III UI 1/3072 1/3072 s +/- 100 ppm
Unit Interval E.48.LV.LV-III UI 1/4915.2 1/4915.2 s +/- 100 ppm
Unit Interval E.60.LV.LV-III UI 1/6144.0 1/6144.0 s +/- 100 ppm
Unit Interval E.96.LV.LV-III UI 1/9830.4 1/9830.4 s +/- 100 ppm

The differential return loss of the receiver shall be better than
-9 dB for 50MHz <= f <2500MHz, and
-9dB + 12*log(f / 2500MHz) dB for 2500MHz <= f <= 7500MHz
The reference impedance for the differential return loss measurement is 100 resistive.
The Common Mode Return Loss is not specified.
Receiver interference tolerance is defined in IEEE 802.3[22] section 72.7.2.1.
6.2.9.5. Equalization and RX Compliance
For HV and LV variant, equalization is allowed by CPRI to overcome data dependent jitter issue. No specific
equalization technique is specified within CPRI.
For LV-II variant, the Equalization performance testing is not independent, but included in jitter tolerance
guided by section 2.4.4 Receiver Interoperability of OIF-CEI02.0 [17].
For LV-III variant, the Equalization performance testing is not independent, but included in receiver
interference tolerance guided by IEEE 802.3 [22] section 72.7.2.1 of 10GBase-KR.
6.2.10. Measurement Procedure
CPRI does not provide means for physical layer conformance testing on chip level or CPRI module level.
The measurement procedures shall be seen as recommendations for the chip manufacturers.
6.2.10.1. Low Voltage Option
Since the Low voltage electrical specification are guided by the XAUI electrical interface specified in Clause
47 of IEEE 802.3-2005 [1], the measurement and test procedures shall be similarly guided by Clause 47. In
addition, the CJPAT test pattern defined in Annex 48A of IEEE 802.3-2005 [1] restricted to lane 0 is specified
as the test pattern for use in eye pattern and jitter measurements. Annex 48B of IEEE 802.3-2005 [1] is
recommended as a reference for additional information on jitter test methods.

6.2.10.2. High Voltage Option
Since the High voltage electrical specification are guided by the 1000Base-CX electrical interface specified in
Clause 39 of IEEE 802.3-2005 [1], the measurement and test procedures shall be similarly guided by Clause
39, with the impedance value 100 instead of 150 . In addition, the CJPAT test pattern defined in Annex
48A of IEEE 802.3-2005 [1] restricted to lane 0 is specified as the test pattern for use in eye pattern and jitter
measurements. Annex 48B of IEEE 802.3-2005 [1] is recommended as a reference for additional information
on jitter test methods.

CPRI
CPRI Specification V4.2 (2010-09-29) 94
6.2.10.3. Low Voltage II Option
Since Low voltage II electrical specification are guided by the CEI-6G-LR electrical interface specified in
Clause 7 of OIF-CEI-02.0[17], the measurement and test procedures shall be similarly guided by Clause 7.
6.2.10.4. Low Voltage III Option
Since Low voltage III electrical specification are guided by the 10GBase-KR electrical interface specified in
Clause 72.7 and Clause 72.8 of IEEE 802.3 [22], the measurement and test procedures shall be similarly
guided by Clause 72.7 and Clause 72.8.
6.3. Networking (Informative)
This chapter is informative and aimed at giving examples of network capabilities of an REC and RE assumed
in CPRI release 2 or higher. It describes the very basic functionality of the REC and RE to support other
topologies than star, e.g. chain, ring or tree topologies.
All functionality described is for informative purpose only and are not mandatory for the REC/RE to
implement. Bi-lateral discussions with a system vendor are necessary for REC/RE requirements.
6.3.1. Concepts
RE
The networking capabilities of an RE supporting CPRI release 2 or higher may differ very much between
implementations. The functionality is therefore described as an interval between a highly capable RE versus
a topology-limited RE. In the following subchapters, the RE functionality is divided into a simple solution
aiming at using a simplified networking functionality in a chain topology as seen in figure 5A and a more
general solution aiming at a chain, tree or ring topology as defined in chapter 2.1..
An RE supporting the general solution is characterized by that it may have several slave ports and several
master ports.
An RE supporting the simple solution is characterized by that it only has one slave port and one master port
which are both using the same line bit rate.

Redundancy
In CPRI release 1, redundancy may exist on hop level by usage of more than one link. In CPRI release 2 or
higher, redundancy may also exist on network level. An RE can be connected to the REC through more than
one logical connection, each logical connection having its own network path.

6.3.2. Reception and Transmission of SAP
CM
by the RE
General solution
SAP
CM
logical connections received on CPRI slave port(s) are switched to CPRI master port(s). The
application layer defines the address table used for switching. It is managed in the REC that has full
knowledge of the topology and all addresses to all REs. The HDLC or Ethernet address can be used to
define a table that maps a CPRI port to an address.
Simple solution
For an RE with one CPRI slave port, all messages from the CPRI slave port are forwarded to the master
port. Messages received on the CPRI master port are forwarded to the CPRI slave port. The forwarding may
be done already at layer 1. The REC must manage the C&M media access in UL (e.g. through a polled
protocol).


CPRI
CPRI Specification V4.2 (2010-09-29) 95
6.3.3. Reception and Transmission of SAP
IQ
by the RE
General solution
SAP
IQ
logical connections received on CPRI slave port(s) are switched to CPRI master port(s). An address
table managed by the application layer defines how SAP
IQ
logical connections shall be switched from one
port to another.
Simple solution
For an RE with only one CPRI slave port, all AxC Containers from the CPRI slave port are forwarded to the
master port. The AxC Containers received on the CPRI master port are forwarded to the CPRI slave port.
The forwarding may be done already at layer 1.

6.3.4. Reception and Distribution of SAP
S
by the RE
General solution
The application layer configures the SAP
S
logical connections, i.e. on which slave port to receive the SAP
S

and to which master ports to distribute the SAP
S
. On the port where SAP
S
is received, the RE must fulfil the
behaviour as described in section 4.2.9 defined for a slave port. On the ports where the SAP
S
is distributed,
the RE must fulfil the behaviour in section 4.2.9 defined for a master port.
If the RE loses the slave port for SAP
S
due to link failure, the SAP
S
is forced to move to another slave port. In
order to support chapter 4.2.9, the whole branch of REs must normally be re-synchronized. The application
layer normally manages the re-synchronisation.
Simple solution
For an RE with only one CPRI slave port, section 4.2.9 shall be fulfilled. The forwarding of SAP
S
to the
master port may be done already on layer 1.

6.3.5. Reception and Transmission of CPRI Layer 1 Signalling by the RE
All layer 1 signalling is per hop basis except for the Reset and the SDI. The LOS, LOF and RAI signals are
read (in each RE) by the application and signalled to the REC via the application layer.
For the layer 1 Reset, see chapter 4.2.7.6.1.
General solution for SDI
The SDI bit received on a CPRI port is switched to other CPRI port(s) depending on their relation to the port
with the SDI set. An address table managed by the application layer defines how the SDI bit shall be
switched from one port to another. It is highly recommended that the SAP
IQ
and SAP
CM
logical connections
are not forwarded from the link where the SDI is set.
Simple solution for SDI
For an RE with only one CPRI slave port, the SDI bit is forwarded to the master port. The forwarding may be
done already at layer 1. It is assumed that the IQ user plane and CM messages are forwarded. A SDI bit
received on a CPRI master port is read by the application and signalled to the REC via the application layer.

6.3.6. Bit Rate Conversion
An RE is allowed to use different bit rates on its CPRI links, e.g. a high-speed slave port and multiple low-
speed master ports.
6.3.7. More than one REC in a radio base station
Up to CPRI release 3 only one REC per base station was considered. Therefore clock/frame synchronization
(sections 3.5 and 4.2.8) and delay calibration (sections 3.6 and 4.2.9) were defined with reference to the
REC. In CPRI release 4 and higher also multiple RECs per base station are considered. In the case of

CPRI
CPRI Specification V4.2 (2010-09-29) 96
multiple RECs the decision which REC is to be taken for clock reference, is assumed to depend on the
individual application. The decision process and the detailed consequences thereof are not described in the
CPRI specification.
In the case of multiple RECs some RECs might also have slave ports. In the latter case section 6.3.8 also
applies.
6.3.8. The REC as a Networking Element
In CPRI release 4.0 and higher, the REC may be used as a networking element (figure 5D and figure 5E in
chapter 2.3). The usage of a networking REC is not fully specified but the following apply.
- Reception and Transmission of SAP
CM
follow chapter 6.3.2.
- Reception and Transmission of SAP
IQ
follow chapter 6.3.3.
- Reception and Distribution of SAP
S
depends on the topology. A REC may follow chapter 6.3.4 and
receive SAP
S
from its slave port and distribute it to its master port(s), but may also distribute its own
SAP
S
to the master port(s).
- Reception and Transmission of CPRI Layer 1 Signalling does not follow chapter 6.3.5. The REC may in
general not do a reset when it receives a reset bit on its slave port. The reception and transmission of all
CPRI Layer1 Signalling is topology dependent.
6.4. E-UTRA sampling rates (Informative)
Typical sampling rates for E-UTRA are derived for the channel bandwidths listed in Table 5.1-1 of 3GPP TS
36.104 [14].
For each channel bandwidth, the total number of sub-carriers in downlink can be computed by the formula:
N
subcarriers
= N
RB
x
RB
sc
N + 1, where
RB
sc
N is equal to 12 (Table 6.2.3-1 of 3GPP TS 36.211 [16])
The size NFFT of the IFFT or FFT operators shall be chosen greater than the number of sub-carriers. Typical
values are listed in Table 20.
The sampling frequency f
S
can be computed using the formula:
f
S
= f x NFFT,
where f the sub-carrier separation is equal to 15kHz (Table 6.12-1 of 3GPP TS 36.211 [16]).

Table 20: typical sampling rates for E-UTRA
Channel bandwidth
(MHz)
1.4 3 5 10 15 20
Number of subcarriers
In downlink
73 181 301 601 901 1201
NFFT 128 256 512 1024 1536 2048
Sampling rate (MHz) 1.92 3.84 7.68 15.36 23.04 30.72
Sampling rate / UMTS
chip rate
1 2 4 6 8

6.5. Scrambling (Normative)
Scrambling is supported depending on the CPRI line rate as shown on Table 21:

Table 21: scrambling support

CPRI
CPRI Specification V4.2 (2010-09-29) 97
Line bit rate Scrambling support Highest available protocol
version number
614.4 Mbit/s Not supported 1
1228.8 Mbit/s Not supported 1
2457.6 Mbit/s Not supported 1
3072.0 Mbit/s Not supported 1
4915.2 Mbit/s Recommended 1: scrambling not supported
2: scrambling supported
6144.0 Mbit/s Recommended 1: scrambling not supported
2: scrambling supported
9830.4 Mbit/s Recommended
1: scrambling not supported
11

2: scrambling supported

6.5.1. Transmitter
The scrambler used is a side stream scrambler as shown in Figure 40. The scrambling sequence is
constructed using the primitive (over GF(2)) polynomial P(X) = 1+X
28
+X
31
.
The scrambling sequence c
i
(i= 0, 1,, 256*16*T-1) is constructed as:
Initial conditions are defined by a 31-bit vector (the seed of the scrambler: c
0
,..c
30
). The choice of the
seed is outside the scope of the CPRI specification. A seed with all bits equal to 0 is not precluded
and allow disabling the scrambling operation.
Recursive definition of subsequent symbols: c
i+31
= c
i
+

c
i+3


modulo 2 for i 0
c
i
bit is the generated bit in time sequence i of the serial pseudorandom code generator (c
0
is the first
outgoing bit).
At each bit period, the shift registers are advanced by one bit and one new bit is generated.
At the beginning of each hyperframe the scrambler state is reset with the seed value (c
0
..c
30
). Hence,
the c
i
sequence period is 256*16*T.

The scrambling sequence generator is followed by a serial to parallel function. The input of this function is
the c
i
sequence. The output is a byte sequence C
n
(n= 0, 1,, 512*T-1) defined by:

C
n
= (c
8n
(LSB), c
8n+1
, ... , c
8n+7
(MSB)) for 0 n < 512*T

Byte sequence C
n
(n= 0, 1,,512*T-1) is defined by following formula to prevent control BYTES #Z.X.Y with
index Y 1 of subchannel Ns=0 (X= 0, 64, 128 and 192) and subchannel Ns=2 (X= 2, 66, 130 and 194) to
be scrambled:

if n {0;1;4T; 4T+1; 128T; 128T+1; 132T; 132T+1; 256T; 256T+1; 260T; 260T+1; 384T; 384T+1; 388T; 388T+1}
C
n
= 0
else
C
n =
C
n
where n= 2*T*X + W*T/8 + Y = 0, 1,..., 512*T-1.

The input of the 8B/10B encoder is the result of a bit wise XOR operation between the byte #Z.X.W.Y
12
and
C
2TX+WT/8+Y



11
At 9830.4 Mbps line bit rate scrambling is strongly recommended.
12
refer to section 4.2.7.1.2 for more details

CPRI
CPRI Specification V4.2 (2010-09-29) 98
The timing relation between the byte #Z.X.W.Y and C
n
is shown in Figure 41.
Serial to Parallel
C2TX+WT/8+Y
(Ns=0 or Ns=2)
& W=0 & Y<=1
8B/10B
encoder
Bit-Wize
XOR
Z.X.W.Y
C2TX+WT/8+Y
0
1
B B XOR
A XOR A
2
3
D D XOR
C XOR C
4
5
F F XOR
E XOR E
6
7
H H XOR
G XOR G
Scrambling Sequence Generator
ci+30 ci+29 ci+28 ci+3 ci+1 ci
c
30
c
29
c
28
c
3
c
1
c
0
Seed vector
XOR
b
a
d
c
i
e
h
f
j
Before scrambling
After scrambling
After scrambling & 8b/10b encoding
ci+2
c
2
0
0
0
0
0
0
0
0
g
ci+31

Figure 40: Scrambling function



CPRI
CPRI Specification V4.2 (2010-09-29) 99
XOR
#Z.128.0.0 #Z.128.0.1 #Z.128.0.2 #Z.128.0.T/8-1 #Z.128.1.0 #Z.128.1.1 #Z.129.15.T/8-1 #Z.130.0.0 #Z.130.0.1 #Z.130.0.2 #Z.130.0.3
#Z.191.15.T/8-1 #Z.191.15.T/8-2 #Z.191.15.T/8-3
Ns = 0 & Y 1
Ns = 2 & Y 1
8B/10B Encoding
XOR XOR XOR
C260T-1
XOR XOR
C260T+1 = 0
XOR
C260T+2
XOR
C260T+3
XOR
C256T+T/8-1
XOR
C256T+T/8
XOR
C256T+T/8+1
XOR XOR XOR
C394T-1
C256T = 0 C256T+1 = 0 C256T+2 C260T = 0
C394T-3 C394T-2
8B/10B Encoding
XOR
#Z.64.0.0 #Z.64.0.1 #Z.64.0.2 #Z.64.0.T/8-1 #Z.64.1.0 #Z.64.1.1 #Z.65.15.T/8-1 #Z.66.0.0 #Z.66.0.1 #Z.66.0.2 #Z.66.0.3
#Z.127.15.T/8-1 #Z.127.15.T/8-2 #Z.127.15.T/8-3
Ns = 0 & Y 1
Ns = 2 & Y 1
8B/10B Encoding
XOR XOR XOR
C132T-1
XOR XOR
C132T+1 = 0
XOR
C132T+2
XOR
C132T+3
XOR
C128T+T/8-1
XOR
C128T+T/8
XOR
C128T+T/8+1
XOR XOR XOR
C256T-1
C128T = 0 C128T+1 = 0 C128T+2 C132T = 0
C256T-3 C256T-2
8B/10B Encoding
XOR
#Z.0.0.0 #Z.0.0.1 #Z.0.0.2 #Z.0.0.T/8-1 #Z.0.1.0 #Z.0.1.1 #Z.1.15.T/8-1 #Z.2.0.0 #Z.2.0.1 #Z.2.0.2 #Z.2.0.3
#Z.63.15.T/8-1 #Z.63.15.T/8-2 #Z.63.15.T/8-3
Ns = 0 & Y 1
Ns = 2 & Y 1
8B/10B Encoding
K28.5 D16.2/D5.6
XOR XOR XOR
C4T-1
XOR XOR
C4T+1 = 0
XOR
C4T+2
XOR
C4T+3
XOR
CT/8-1
XOR
CT/8
XOR
CT/8+1
XOR XOR XOR
C128T-1
C0 = 0 C1 = 0 C2 C4T = 0
C128T-3 C128T-2
8B/10B Encoding


CPRI
CPRI Specification V4.2 (2010-09-29) 100


Figure 41: Scrambling of bytes #Z.X.W.Y in hyperframe Z

6.5.2. Receiver
A receiver supporting protocol version 2 shall be capable of receiving data scrambled by the scrambling
function described in section 6.5.1 for any seed value.
The receiver shall use at least 31 bits in the control BYTES #Z.0.2 to #Z.0.(T/8-1) to retrieve the scrambling
sequence of the transmitter in order to generate the descrambling sequence.
Once the above operation is achieved, the receiver shall periodically check the descrambling sequence with
the incoming data, by sampling at least 31 bits of the descrambled control BYTES #Z.0.2, to #Z.0.(T/8-1)
known to be 50h (see 4.2.10.3.1)


CPRI
CPRI Specification V4.2 (2010-09-29) 101
7. List of Abbreviations
AC Alternating Current
A/D Analogue/Digital
ANSI American National Standardization Institute
AxC Antenna-carrier
BER Bit Error Ratio
BFN Node B Frame Number
C Control
ceil() The function ceil returns the smallest integer greater than or equal to the argument.
CP Cyclic Prefix
C&M Control and Management
CPRI Common Public Radio Interface
D/A Digital/Analogue
DA Destination Address
DL Downlink
ESD End-of-Stream-Delimiter
E-UTRA Evolved Universal Terrestrial Radio Access
f
C
Chip Rate of UTRA-FDD = 3.84MHz
FCS Frame Check Sequence
FDD Frequency Division Duplex
FFT Fast Fourier Transform
floor() The function floor returns the greatest integer less than or equal to the argument.
f
S
Sampling rate
GF Galois Field
GPS Global Positioning System
HDLC High-level Data Link Control
HFN Hyper Frame Number
HV High Voltage
I In-Phase
IEC International Electrotechnical Commission
IEEE Institute of Electrical and Electronics Engineers
IFFT Inverse Fast Fourier Transform
Iub Interface between Radio Network Controller and UMTS radio base station (NodeB)
LCM Least Common Multiple
LLC Logical Link Control
Ln Length
LOF Loss of Frame
LOS Loss of Signal
LSB Least Significant Bit

CPRI
CPRI Specification V4.2 (2010-09-29) 102
LV Low Voltage
LVDS Low Voltage Differential Signal
M Management
MAC Media Access Control
MIMO Multiple Input, Multiple Output
MSB Most Significant Bit
N
RB
Number of resource blocks in an E-UTRA cell
RB
sc
N Resource block size in the frequency domain, expressed as a number of subcarriers
N/A Not Applicable
PAD Padding
PCS Physical Coding Sublayer
PDU Protocol Data Unit
PHY Physical Layer
PLL Phase Locked Loop
PMA Physical Medium Attachment
Q Quadrature
RAI Remote Alarm Indication
RE Radio Equipment
REC Radio Equipment Control
RF Radio Frequency
RRC Root Raised Cosine
Rx Receive
SA Source Address
SAP Service Access Point
SDI SAP Defect Indication
SDU Service Data Unit
SERDES SerializerDeserializer
SFD Start-of-Frame Delimiter
SFP Small Form-factor Pluggable
SSD Start-of-Stream Delimiter
T Number of bits per (control) word in a CPRI basic frame as defined in section 4.2.7.1
T
C
CPRI basic frame length = UTRA FDD Chip period = 1/3.84MHz
T
F
WiMAX frame length
TP Test Point
TS Technical Specification
Tx Transmit
UE User Equipment
UL Uplink
UTRA Universal Terrestrial Radio Access (3GPP)
UTRAN Universal Terrestrial Radio Access Network (3GPP)

CPRI
CPRI Specification V4.2 (2010-09-29) 103
UMTS Universal Mobile Telecommunication System
Uu UMTS air interface
WiMAX Worldwide Interoperability for Microwave Access
XAUI 10 Gigabit Attachment Unit Interface
Z.X.W.Y Byte Index (byte number Y, word number W, basic frame number X, hyperframe number Z)
#Z.X.W.Y Content of byte with index Z.X.W.Y
Z.X.Y Short form of BYTE Index, for control BYTES only (word number W = 0)
#Z.X.Y Content of control BYTE with index Z.X.Y
3GPP 3
rd
Generation Partnership Project


CPRI
CPRI Specification V4.2 (2010-09-29) 104
8. References
[1] IEEE Std 802.3-2005: "Part 3: Carrier sense multiple access with collision detection (CSMA/CD)
access method and physical layer specifications",12 December 2005.
[2] IEEE Std 802.3ae-2002 Part 3: Carrier sense multiple access with collision detection (CSMA/CD)
access method and physical layer specifications Amendment: Media Access Control (MAC)
Parameters, Physical Layers, and Management Parameters for 10 Gb/s Operation, March 2002.
[3] ISO/IEC 14165-115 Information Technology Fibre Channel Part 115 Physical Interface (FC-PI),
February 2006.
[4] IEC 60793-2-10 (2002-3) Part 2-10: Product specifications Sectional specification for category A1
multimode fibres, March 2002.
[5] IEC 60793-2-50 (2002-1) Part 2-50: Product specifications Sectional specification for class B
single-mode fibres, January 2002.
[6] Infiniband Trade Association: Infiniband Architecture, Rel. 1.1, Vol. 2, November 2002.
[7] ANSI: ANSI-TIA-644, January 2001.
[8] 3GPP TS 25.104: Base Station (BS) radio transmission and reception (FDD), Release 9, V 9.3.0,
March 2010.
[9] 3GPP TS 25.133: Requirements for support of radio resource management (FDD), Release 9, V
9.3.0, March 2010.
[10] ISO/IEC: Information technology Telecommunications and information exchange between
systems High-level data link control (HDLC) procedures. International Standard ISO/IEC 13239,
3
rd
edition, Reference number: ISO/IEC 13239:2002(E), 2002-07-15.
[11] WiMAX Forum WMF-T23-001-R015v01, WiMAX Forum Mobile System Profile, Release 1.5
Common Part (2009-08-01) and WiMAX Forum WMF-T23-003-R015v01, WiMAX Forum Mobile
System Profile Specification, Release 1.5 FDD Specific Part, (2009-08-01) and WiMAX Forum
WMF-T23-002-R015v01, WiMAX Forum Mobile System Profile Specification, Release 1.5 TDD
Specific Part, (2009-08-01)
[12] IEEE Std 802.16e-2005 and IEEE 802.16-2004/Cor1-2005, IEEE, New York, USA, 28 February
2006
[13] IEEE Std 802.16-2009, IEEE Standard for Local and metropolitan area networks - Part 16: Air
Interface for Broadband Wireless Access Systems
[14] 3GPP TS 36.104: Evolved Universal Terrestrial Radio Access (E-UTRA), Base Station (BS) radio
transmission and reception, Release 9, V 9.3.0, March 2010.
[15] 3GPP TS 36.133: Evolved Universal Terrestrial Radio Access (E-UTRA), Requirements for support
of radio resource management, Release 9, V9.3.0, March 2010.
[16] 3GPP TS 36.211: Evolved Universal Terrestrial Radio Access (E-UTRA), Physical Channel and
Modulation, Release 9, V9.1.0, March 2010.
[17] Electrical I/O (CEI) - Electrical and Jitter Interoperability agreements for 6G+ bps and 11G+ bps
I/O, IA # OIF-CEI-02.0, 28th February 2005
[18] INCITS (ANSI) Revision 8, T11/08-138v1 Fibre channel Physical Interface-4 (FC-PI-4), May 21
st

2008.
[19] INF-8074i - Specification for SFP (Small Formfactor Pluggable) Transceiver, Revision 1.0, May 12
th

2001.
[20] SFF-8431 - Specification for Enhanced 8.5 and 10 Gigabit Small Form Factor Pluggable Module
"SFP+", Revision 3.2, Nov 12
th
2008.
[21] SFF-8083 - Specification for 0.8mm SFP+ Compliant Card Edge Connector, Revision 2.0, Oct 17
th

2008.
[22] IEEE Std 802.3-2008 IEEE, New York, USA, 26th December 2008

CPRI
CPRI Specification V4.2 (2010-09-29) 105
9. History
Version Date Description
V 1.0 2003-09-30 First complete CPRI specification
V 1.1 2004-05-10 Editorial corrections.
Section 3: Clarification of input requirements for CPRI.
Section 4.2.7.5: An additional sequence K28.5 + D5.6 (defined in
the 8B/10B standard as /I1/) is allowed for the use as control sync
word to enable usage of existing SERDES devices.
Section 4.5.3.7: Editorial correction in subsection RE actions to
align the text with Figure 30.
Section 5.1.4: Update of specification release version.
Section 5.2: Clarification of CPRI implementation compliancy.
V 1.2 2004-07-15 Sections 4.2.2 to 4.2.4: Recommendation of a low voltage (CX
based) and a high voltage (XAUI based) electrical interface.
Addition of Section 6.2.
Editorial changes and abbreviation addition.
V 1.3 2004-10-01
Major editorial correction in Section 4.5.4.4 and Section 4.5.4.12:
Exchange of BYTE index Z.64.0 with Z.66.0
Exchange of BYTE index Z.192.0 with Z.194.0

CPRI
CPRI Specification V4.2 (2010-09-29) 106
V 2.0 2004-10-01 Introduction of the CPRI networking feature resulting in the following list of
detailed modifications:
Chapter 1:
Clarification of the CPRI scope (layers 1 + 2).
Clarification of the support mechanisms for redundancy.
Section 2.1:
Additional definitions for node, link, passive link, hop, multi-hop,
logical connection, master port and slave port.
Section 2.2:
Update of system architecture introducing links between REs.
Section 2.3:
Addition of chain, tree and ring topologies.
Section 2.4:
Addition of the Section 2.4.2 on the CPRI control functionality.
Chapter 3:
Adaptation of the requirements to the networking nomenclature.
Scope of each requirement has been added.
Section 3.3:
Addition of chain, tree and ring topologies.
New requirements for no. of hops and ports have been added.
Section 3.5.1:
Requirement of clock traceability for RE slave ports.
Section 3.5.2:
Transparent forwarding of frame timing information.
Section 3.5.3:
Renaming of section to link timing accuracy.
Clarification of requirement.
Section 3.6:
Introduction of subsection 3.6.1 covering the round trip cable delay
measurement requirements for the link.
Addition of subsection 3.6.2 on the round trip delay measurement
requirements for a multi-hop connection.
Section 3.9.2:
Requirement on the auto-detection of REC data flow on slave ports
has been added.


CPRI
CPRI Specification V4.2 (2010-09-29) 107
Section 4.2.7.6.1:
Forwarding of reset bit has been added.
Section 4.2.7.6.2:
Clarification has been added that the filtering applies to reset as
well as reset acknowledgement.
Section 4.2.8:
Redefinition of synchronization and timing source.
Section 4.2.9:
Renaming of section heading
Multi-hop case and multiple slave ports case are considered.
New reference points RB1-4 were defined. Figure 24A was added.
Timing relations of multi-hop configuration were defined. Figure 25A
was added.
Section 4.5:
REC is replaced by master port.
RE is replaced by slave port.
The terms Uplink and Downlink are replaced to avoid confusion
in case of a ring topology.
The text of the sections defining transitions 1 and 11 is updated.
Section 5.1.4:
Update of specification release version.
Annex 6.1:
Delay calibration example for multi-hop configuration has been
added.
Annex 6.3:
Addition of an Annex called Networking aiming at giving examples
of network capabilities of an REC and RE assumed in CPRI version
2.0.
Section 7:
Update of list of abbreviations.
Section 9:
Update of history.

In addition, minor editorial corrections have been made.
V 2.1 2006-03-31 Chapters 3 and 8:
Update of the requirement no. R-1 as well as of References [8] and
[9] to 3GPP UTRA FDD, Release 6, December 2005
Minor editorial correction in Section 4.2.7.5:
Table 9: Change X to 0 #Z.0.0 #Z.0.1 #Z.0.2 #Z.0.3


CPRI
CPRI Specification V4.2 (2010-09-29) 108
V 3.0 2006-10-20 Introduction of WiMAX resulting in the following list of detailed modifications:
Chapter 2:
New definitions/nomenclature, system architecture, and functional
split for WiMAX added
Chapter 3:
Update of requirements R-1, R-5,, R-12, R-19,, R-21A, R-30
New requirements for WiMAX: R-4F, R-11A, R-12A, R-20A
Section 4.2.7.2:
WiMAX IQ mapping added including new subsections 4.2.7.2.4
through 4.2.7.2.7
New subsection 4.2.7.2.8 for WiMAX TDD/FDD added
Section 4.2.8 and section 6.1:
Synchronisation and timing for WiMAX specified
Section 5.1.4:
Protocol version number for CPRI V3.0 specified
Introduction of line bit rate option 4 (3072.0Mbit/s) resulting in the following
list of detailed modifications:
Section 4.2.1:
New line bit rate option 4 listed
Section 4.2.2:
Physical layer modes for line bit rate option 4 added
Section 4.2.7.1:
Basic frame structure for line bit rate option 4 added
Section 4.2.7.3:
Line bit rate option 4 added to hyperframe structure
Section 4.2.7.5:
Synchronization control word for line bit rate option 4 specified
Section 4.2.7.6 and section 4.2.7.7.1:
New configurations of slow C&M channel for line bit rate option 4
added
Section 4.2.7.7.2:
New configurations of fast C&M channel for line bit rate option 4
added
Section 6.2.:
Physical layer specification for line bit rate option 4 added
Update of Chapters 7, 8, and 9.
In addition, minor editorial corrections have been made.

CPRI
CPRI Specification V4.2 (2010-09-29) 109
V4.0 2008-06-30 Introduction of LTE & MIMO resulting in the following list of detailed
modifications
Chapter 2:
New definitions/nomenclature, system architecture, and functional
split for E-UTRA added
Chapter 3:
Update of requirements R-1, R-11A, R-12A, R-19, R-20, R-20A, R-
21, R-21A, R-26
Section 4.2.7.2.:
E-UTRA IQ-mapping added
Section 4.2.8 and section 6.1:
Synchronisation and timing for E-UTRA specified
Figure 31 modified
Section 5.1.4:
Protocol version number for CPRI V4.0 specified
Chapter 6:
New informative section 6.4 E-UTRA sampling rates added
Chapter 7:
Update of the abbreviation list
Chapter 8:
Update of Reference list
Introduction of multiple REC topologies resulting in the following list of
detailed modifications:
Chapter 1:
Scope of the specification modified in order to also cover multiple
REC topologies
Section 2.1:
Basic nomenclatures modified in order to also cover multiple REC
topologies
Section 2.3:
Multiple REC configurations added / new figures added showing
multiple REC topologies
Section 4.1, 4.2.9, 6.1:
Footnotes added
Section 6.3:
New subsections 6.3.7 and 6.3.8 for multiple REC topologies added

CPRI
CPRI Specification V4.2 (2010-09-29) 110
Addition of oversampling ratio 2 for UTRA FDD Downlink resulting in the
following list of detailed modifications:
Section 3.4.2:
Notes updated
Section 4.2.7.2.2:
Modification of Table 5 and introduction of new Table 5A
Figure 11: figure caption modified
In addition, the following modifications were done:
Sections 3.5.3 and 3.5.4:
Addition of a note on the scope of TX delay being link (below R-19
and R-20 respectively)
Section 3.5.3:
Improved wording of Link Timing Accuracy"
Section 4.2.2, 4.2.3, 4.2.4, 8
Replacement of references to INCITS 352 by ISO/IEC 14165-115
Sections 4.2.2 4.2.5, 4.2.7.1.2, 4.4, 6.2, 8
Replacement of references to IEEE 802.3 2002 / IEEE 802.3ae-
2002 by IEEE Std 802.3-2005
Section 4.2.7.7.3:
Allowance of simple RE with no or simple C&M-link (use of non-
zero C&M-channel is now recommended rather than mandatory)
Section 4.5.3.2:
Slave port actions modified for improved LOS/LOF handling
Section 9:
Update of history

In addition, minor editorial corrections have been made.

CPRI
CPRI Specification V4.2 (2010-09-29) 111
V4.1 2009-02-18 Introduction of higher line rates (x8 and x10) for CPRI resulting in the
following list of detailed modifications:
a) Physical layer characteristics:
Section 4.2.1:
New CPRI line bit rate options 5&6 introduced (8x & 10x)
Section 4.2.2:
Table 2: New CPRI physical layer modes 4915.2 Mbps & 6144
Mbps included
Figure 6A: new LV-II variant included
Sections 6.2, 6.2.1, 6.2.4, 6.2.6 and 6.2.7:
New LV-II variant adopted
Sections 6.2.8.3 & 6.2.9.3
New sections defining electrical Tx- & Rx-characteristics of LV-II
Section 6.2.8.3:
Modified w.r.t. TX-compliance
LV-II variant included
Section 6.2.9.4:
New section for Equalization and RX-compliance
Section 4.2.24.2.4 & 8:
New reference [17] for OIF-CEI added
References to Fibre Channel Physical Interface-4, SFP and SFP+
introduced
b) Introduction of data scrambling:
Section 3.9.2:
New requirement R-31A (Autonegotiation of Scrambling)
Section 4.2.7.1.2:
New title: Transmission Sequence and Scrambling
Scrambling impact on transmission sequence defined
Section 4.2.7.6:
New protocol version #Z.2.0 = 2 introduced in table 10
Section 4.2.10.3.1:
New Figure 26A (LOF and HFNSYNC detection with scrambling
enabled)



CPRI
CPRI Specification V4.2 (2010-09-29) 112
Section 5.1.4:
New specification release 4.1 added to table 15
New protocol version number 2 introduced for this specification
release in table 15
Section 6.5:
New normative scrambling section
c) Impact on frame structure and HDLC-rate:
Section 4.2.7:
Generic basic frame structure introduced (Figure 9B)
Table 6, 9 and 12 extended to also cover x8 & x10 line rates
Table 10 and 11 extended to also cover HDLC bit rate negotiation
on higher layers
New Figure 22B
Sections 4.5.3.4 & 4.5.3.5:
HDLC rate negotiation included

In addition the following modifications were done:
Sections 3.1 and 8
New versions of the 3GPP and WiMAX Forum specifications
adopted
Section 3.5.3
Time alignment between requirement between branches now
defined in 3GPP TS 36.104
Footnote eliminated
Section 3.6.1 & 3.6.2:
Correction of delay calibration description
Section 3.5.3 & 3.5.4:
Text improvements
Section 4 & 6:
Consistent usage of Z.X.(W).Y and #Z.X.(W).Y




CPRI
CPRI Specification V4.2 (2010-09-29) 113
V4.2 2010-09-29 Introduction of higher line rate (16x) for CPRI resulting in the following list of
detailed modifications:
a) Physical layer characteristics:
Section 4.2.1:
New CPRI line bit rate option 7 introduced (16x)
Section 4.2.2:
Table 2: New CPRI physical layer mode 9830.4 Mbps included
Figure 6A: new LV-III variant included
Section 4.2.2 & 8:
New reference [22] for IEEE Std 802.3-2008 added
Section 5.1.4:
New version 4.2 in Table 15
Sections 6.2, 6.2.1, 6.2.4, 6.2.6 and 6.2.7:
New LV-III variant adopted
Sections 6.2.8.4 & 6.2.9.4
New sections defining electrical Tx- & Rx-characteristics of LV-III
Section 6.2.8.5 (former section 6.2.8.4):
LV-III variant included
Section 6.2.9.5 (former section 6.2.9.4):
LV-III variant included
Section 6.2.10:
New section Low voltage III option
Section 6.5:
Table 21: new line bit rate 9830.4 Mbps
b) Impact on frame structure and HDLC-rate:
Section 4.2.7:
Table 3, 6, 9, 11 and 12 extended to also cover 16x line rate
In addition the following modifications were done:
Section 1 and cover page:
Removal of Nortel reference
Sections 3.1, 3.5.4, 3.6.1, 3.6.2 and 8:
Update of 3GPP- and WiMAX-references
Removal of footnotes 3&4
Section 4.2.7.5:
Missing #Z.0.5 added to Table 9
Section 4.2.7.2.4
Figure 13A corrected (s-3)