!!CONTENT!!
1) Introduction to ATM on Multiservice Switch networks
Overview of ATM network architecture
ATM on Multiservice Switch networks
ATM function processors
Inverse multiplexing for ATM
Multiservice Switch trunks over ATM
Frame relay over ATM
2) ATM connections and signaling protocols
Connections
ATM signaling protocols
3) Traffic management for ATM
Traffic contract
Traffic shaping
Traffic scheduling
Memory management
Dynamic bandwidth management
4) ATM TRANSMISSION
Introduction
Logical Connection
5) ATM Multiplexing
6) ATM & Frame Relay
Wide area network internetworking
Service internetworking
7) Summery of ATM
ATM Characteristics
Advantages
Disadvantages
Introduction to ATM on Multiservice Switch
networks
Asynchronous transfer mode (ATM) is a high bandwidth, low-delay,
connection-oriented switching and multiplexing technique. ATM
technology is based on the switching of small fixed-length packets of
data called cells. In ATM, all data is transferred in 53-byte cells. Each
cell has a 5-byte header that identifies the cells route through the
network and 48 bytes that contain user data (called the payload).
Overview of ATM network architecture
The adaptation layer is the outer layer of the ATM network. This
layer includes enterprise switches and service access multiplexers.
These devices support the service interfaces and adapt legacy
protocols, such as frame relay, to ATM. The Passport 6400 switch is
an ATM access device and can exist either on the customer premises
or in the service providers central office.
The access layer is the middle layer of the ATM network. Nortel
Multiservice Switch 7400 devices are access concentration nodes
or edge nodes. These nodes receive cells from the nodes in the
adaptation layer and concentrate these cells into the ATM
backbone layer.
The ATM backbone is the central layer in the ATM network. A
Nortel Multiservice Switch 15000 is responsible for routing traffic
through the network at high speeds.
ATM network architecture
ATM on Multiservice Switch networks
ATM on Multiservice Switch networks is based on standards
developed by the ATM Forum and the International
Telecommunication Union (ITU). Because ATM uses small cells to
transfer data and isochronal timing, Multiservice Switch devices can
support a wide range of audio, video, image, and data
communications requirements.
ATM uses a type of time division multiplexing that transmits different
applications (voice, data, or video) over the same connection. With
ATM technology, Multiservice Switch nodes can allocate bandwidth
on demand based on each applications bandwidth and quality of
service (QOS) requirements.
Multiservice Switch nodes support the following ATM services to
allow service providers to use other forms of traffic in their
networks:-
Inverse multiplexing for ATM.
AAL1 circuit emulation service.
Multiservice Switch trunks over ATM.
Frame relay over ATM.
ATM multiprotocol encapsulation service.
Multiservice Switch nodes support the following advanced traffic
management capabilities to allow service providers to optimize
network performance:-
connection admission control
bandwidth pool management
resource management
queue management
congestion control
traffic shaping and policing
ATM function processors
Nortel Networks defines ATM function processors (FPs) by the type of
application specific integrated circuit (ASIC) and by the type of field
programmable gate array (FPGA) on the card:
APC-based FPs use the ATM Port Controller (APC) ASIC, the Multiservice
Switch queue controller (PQC) ASIC, and the Queue Relay Device (QRD)
FPGA to provide enhanced functionality.
AQM-based FPs use the Multiservice Switch queue controller (PQC) and
ATM queue manager (AQM) ASICs to provide enhanced functionality.
However, the 1-port OC-12 FPs use the PQC and the ATM queue
scheduler (AQS) FPGA to provide OC-12 capacity.
AQS-based FPsNortel Networks name for the Multiservice Switch
ATM FP that provides high performance ATM adaptations for IP, frame
relay, and existing networking applications.
CQC-based FPa type of Multiservice Switch ATM FP that uses a cell
queue controller (CQC) ASIC. This type of FP is the original ATM FP on
Multiservice Switch devices.
The Multiservice Access FPs use the Multiservice Switch node queue
controller (PQC) and ATM queue manager (AQM) ASICs to provide
enhanced functionality.
The table Function processor ASICs maps the FP to the type of ASIC
and Multiservice Switch series.
Inverse multiplexing for ATM
Inverse multiplexing for ATM (IMA) is a feature that supports the
transparent transmission of ATM cells over a combination of multiple
DS1/E1 links (an IMA link group). The link group uses the inverse
multiplexing process to transmit a single stream of ATM layer traffic
across multiple links. IMA then combines the traffic back into the original
cell sequence at the remote end.
IMA on Nortel Multiservice Switch nodes supports the use of
synchronized and non-synchronized links within an IMA link group. Use
an IMA link group within a Multiservice Switch network to access an
external ATM network. IMA is available to both private and public user-
to-network (UNI) or network-to-network(NNI) interfaces. The Multiservice
Switch networks traffic management capabilities also apply to ATM
connections served by IMA link groups.
Multiservice Switch trunks over ATM
The Multiservice Switch node-to-Multiservice Switch node links are
called Multiservice Switch trunks (trunks). A trunk is a point-to-point or
logical connection between two Multiservice Switch nodes over which
Multiservice Switch proprietary routing protocols run. Multiservice Switch
trunks over ATM implement an unacknowledged protocol and support
ATM connections for node-to-node connectivity.
Trunks over ATM allow all non-ATM Multiservice Switch services and
DPN-100 traffic to travel transparently over ATM while allowing the
transport of all existing services. This transport mechanism also enables
new service offerings to be carried prior to standardization of future
adaptation protocols.
Frame relay over ATM
The frame relay to ATM (FR-ATM) interworking service for Nortel
Multiservice Switch nodes allows the carriage of frame relay traffic over
an ATM networking and transport infrastructure. It maps frame relay
permanent virtual connections (PVC) to and from ATM PVCs to provide
connectivity between frame relay customer premises equipment (CPE)
and ATM CPE. The FR-ATM interworking function supports the application of
either the FRF.8 standard to support service interworking (SIWF), or the FRF.5
standard to support network interworking (NIWF).
ATM connections and signaling protocols
Connections
The Multiservice Switch ATM connection provides end-to-end
information transfer capability to access points.
A virtual channel connection (VCC) is an ATM connection where
switching is performed on the virtual path identifier (VPI) and virtual
channel identifier (VCI) fields of each cell. A VPI is a field in the ATM cell
header that indicates the virtual path for routing the cell. A VCI is a
unique numerical tag, as defined by a 16-bit field in the ATM cell header
that identifies the virtual channel over which the cell is to travel. A virtual
path connection is an ATM connection where routing is performed only
on the VPI field of each cell.
Multiservice Switch nodes support four ATM connection types:-
Permanent virtual channels and paths.
Switched connections.
Virtual path terminators.
Permanent virtual channels and paths
A permanent virtual connection can be either a virtual path (PVP) or
virtual channel (PVC). A permanent connection has a predefined static
route that provides a permanently configured connection between the
customer premise equipment and the ATM networks. Permanent
connections are set up using predetermined user requirements for
bandwidth and the duration of the connection. Once configured,
permanent connections remain set up even when they are not in use.
Switched connections
A switched connection supports the same functionality as permanent
connections and provides dynamic establishment at each node along
the connection route. The connection route is automatically selected.
Switched connections do not require configuration, but network nodes
must be configured for ATM routing.
Point-to-point connections provide the capability for bi-directional unicast
data communication. Data traffic moves simultaneously in both
directions, and allows for the possibility of the two streams of traffic
requiring different bandwidths.
Point-to-multipoint connections provide the capability for unidirectional
multicast data communications using VCC-based switched connections.
A point-to-multipoint connection is a collection of associated ATM virtual
channels with associated end point nodes. Data traffic moves in one
direction only, from the single source to the multiple destinations.
Virtual path terminators
A virtual path terminator (VPT) is an ATM network entity that unbundles
the virtual channels of a virtual path for independent processing of each
virtual channel. There are two types of VPTs, basic and standard, that
serve as connection end points of a virtual path connection.
Basic VPT
For basic VPTs, independent virtual connections configured within a basic
VPT are equivalent from a datapath and traffic management perspective. Basic
VPTs support the following capabilities:
virtual path operations and maintenance functionality (end-to-end loopbacks
and virtual path to virtual channel fault interworking).
spooled VCC statistics and accounting records.
optional VPT connection admission control (CAC).
Standard VPT
Standard VPTs provide traffic management capabilities at both the virtual path
and virtual channel levels. Standard VPTs can dynamically share bandwidth
among virtual paths and virtual channels within a given virtual path. In addition
to the features supported by basic VPTs, standard VPTs also support the
following capabilities:
real-time statistics (operationally viewable as opposed to spooled)
implemented at the VPT level.
virtual path shaping and weighted fair queuing (WFQ).
multiplexing of multiple virtual channel service classes.
ATM signaling protocols
The Multiservice Switch ATM signaling interface is the connection
between two network nodes or entities. Each entity is either customer
premises equipment (CPE), another Multiservice Switch node, or an
external ATM network. The ATM interface layer provides the direct
association between the interface and the physical port and, by
extension, the physical link hard-wired to the port. Each physical ATM
link or inverse multiplexing for ATM (IMA) link group has one ATM
interface at each end. The interface, and the networking and protocol
associated with that interface, control connections on the ATM link.
Multiservice Switch nodes support the following ATM signalling interface
types:
User-to-network interface.
Interim interswitch signalling protocol interface.
ATM inter-network interface.
Private network-to-network interface.
Virtual interfaces.
Traffic management for ATM
The communication needs of these applications translate into a set of traffic
characteristics based on the required quality of service (QOS) classes and
the traffic descriptor types. Multiservice Switch provides traffic
management functions that ensure that the QOS objectives for each
subscriber are met. Multiservice Switch traffic management strategies also
optimize the service providers use of network resources so that the service
offering is cost effective.
Most traffic management functions apply to both PVC connections and
switched connections (including SPVCs), although some features cannot be
selected for individual VCCs.
The figure Overview of application points for traffic management controls
shows where these controls apply along the data path from link ingress to
link egress.
Traffic contract
A traffic contract defines the traffic characteristics of a connection in the
ATM network. The service provider and the subscriber agree on the level of
service that each connection must support. The service provider defines
the service requirements for each subscriber according to the applications
that the network must support. This definition requires translating the
communications needs of these applications into a set of traffic
characteristics. Nortel Multiservice Switch.
ATM service categories
constant bit rate (CBR)The CBR service category supports
real time applications that require tightly constrained delay
and delay variation. These applications include voice and
video applications. The consistent availability of a fixed
quantity of bandwidth is appropriate for CBR service. Cells
that the network delays beyond the value for CTD are of
significantly reduced value to the application.
real time variable bit rate (RT-VBR)The RT-VBR service
category also supports for real time applications that require
tightly constrained delay and delay variation. These
applications include voice and video applications. The
network expects sources to transmit at a rate that varies
with time. Standards describe this source as bursty. Cells
that the network delays beyond the value for CTD are of
significantly reduced value to the application. Real-time VBR
service can support statistical multiplexing of real time
sources.
non-real time variable bit rate (NRT-VBR)The NRT-VBR
service category supports non-real time applications that
have bursty traffic characteristics and that standards
characterize in terms of a PCR, SCR and MBS. For cells that
the network transfers within the traffic contract, the
application expects a low cell loss ratio. For all connections,
the application expects a limit to the mean cell transfer
delay. Non-real time VBR service support statistical
multiplexing of connections.
Broadband bearer capability parameters
The Multiservice Switch devices support the following
broadband bearer capability (BBC) parameters for switched
connections:
bearer class
transfer capability
clipping
best effort
frame discard (in the transmit and receive directions)
Connection traffic descriptors
The Multiservice Switch devices support the following
connection traffic descriptors:
peak cell rate (PCR)
sustained cell rate (SCR)
maximum burst size (MBS)
cell delay variation tolerance (CDVT)
Traffic shaping
The Multiservice Switch devices traffic shaping strategies ensure that
transmitted traffic conforms to subscribed traffic parameters by regulating
the emission interval of cells transmitted on the link. By pacing the
transmission of cells, the service provider avoids cell discard at points
where traffic policing applies. A devices traffic shaping strategies also
allow the network manager to balance the quality of service requirements
against network costs. Multiservice Switch supports traffic shaping in all
supported service categories.
Traffic scheduling
The service provider requires appropriate traffic scheduling policies to
meet the contracted ATM service category for each node in the network.
On Nortel Multiservice Switch device ATM function processors, there are
queues at both the incoming and outgoing link sides and both to and
from the bus. The devices scheduling policy is based on a system of
emission and discard priorities. This system applies different urgency
priorities to different traffic types so that each application can use the
correct ATM service category to meet the urgency and value
requirements.
Memory management
The Multiservice Switch devices support resource control mechanisms
that manage the queue and memory resources. These resources
monitor and control the link, processor, bus, and memory resources on
the function processor. The devices resource control mechanisms
permit you to configure ATM queue management connection pools and
frame connection resources for sub connections, and Multiservice
Switch dynamic packet routing system (DPRS) connections.
Dynamic bandwidth management
If the bandwidth reduces so that the switch can no longer support
existing connection, Nortel Multiservice Switch devices identify lower
priority connections that can terminate or can support bandwidth
reduction. This capability reduces potential congestion at the ATM
interface and maintains.
ATM TRANSMISSION
Introduction
ATM is designed to be a connection-oriented technology. Instead of
broadcasting transmissions for all stations on the network to receive, a
specific virtual circuit is set up between two end devices. A connection
between the two devices is established, similar to a phone call, before
any transmission occurs.
The ATM cells are identified with a particular virtual circuit and are
delivered only to devices on that virtual circuit.
Logical connections
The virtual circuits used in the ATM environment define the logical
channels through which communications occur. As was mentioned, two
types of virtual circuits exist, Permanent Virtual Circuits (PVCs) and
Switched Virtual Circuits (SVCs). It is promised that ATM will support
both. However, at the present time there is little support for SVCs. They
will be required to be able to establish connections between individual
stations, with sessions lasting only as long as needed. It is expected that
the PVC method will prevail until reliable congestion management
systems are developed.
ATM logical channels are defined by two components. The first is the
Virtual Channel(VC) which is the actual logical connection between
devices. The second is the Virtual Path (VP), which is a bundle of
virtual channels.
Also in the ATM cell header is the Virtual Channel Identifier (VCI) which
identifies the virtual channel within the virtual path. The channels within
a VPI are identified relative to the virtual path.
ATM multiplexing
ATM transmission is based on the statistical modelling of various type of
data. This modelling is done in the ATM network device. An ATM switch
collects the information from many attached devices. This information is
multiplexed into a single stream of cells, which is then sent to the
destination ATM switch. At the receiving end, the stream of cells is split
back into their original streams. Because of its multiplexing of discrete
data streams, ATM is sometimes referred to as Asynchronous Time
Division (ATD).
ATM and Frame Relay
As was mentioned, the ATM Forum is working in conjunction with the
Frame Relay Forum to provide specifications for implementing ATM and
Frame Relay together. The emphasis is on co-existence. Frame Relay is
a technology in widespread use today while ATM is seen as a likely
choice for a new network capable of supporting Frame Relay as one of
several interface access technologies.
Both ATM and Frame Relay are based on packet-switching. A major
difference is speed potential. Frame Relays maximum speed is 1024
Mbps (1.024 Gbps) while ATM speeds of 2.4 Gbps are already
available, with even higher speeds expected.
At this time, the ATM and Frame Relay Forums are close to completing
specifications in two areas. These are the following:
Wide-Area Network Internetworking.
This would allow two Frame Relay networks to communicate over an
ATM backbone. There would be no need to change the Frame Relay
networks but they would be able to take advantage of ATMs high
bandwidth.
The Frame Relay emulation, or translation, would take place in customer
premises equipment, connecting to either a circuit providers or a private
ATM network.
Summary of ATM
ATM characteristics
ATM technology can be described by the following characteristics:
ATM is a cell-relay technology, transmitting data packets of a fixed
size called cells.
Connections between communicating devices in an ATM network
are made by establishing virtual circuits between them.
ATM technology allows the amount of available bandwidth to be
predicted and guaranteed.
ATM is a switch-based technology, making it scale able. As traffic
loads increase, additional switches can be added to the network.
Data transfer rates are very efficient once the connection has been
established.
Advantages
Following is a summary of some of the advantages associated with
ATM:
Error handling on a link-by-link basis is eliminated, thereby increasing
throughput.
Transmissions occur in a connection-oriented mode, with provisions for
any-to-any connections.
The amount of header information is less than that found in many other
technologies.
ATM cells are relatively small and are consistent in size. This reduces
the type of delays at switches that would otherwise be caused by larger
and/or variable-sized packets.
Disadvantages
Following is a summary of some of the disadvantages associated with
ATM:
Lack of standards in key areas.
Carriers are seen as being slow to provide wide-area ATM services.
New technologies are often viewed as being unstable. Many potential
ATM users are waiting for second and third generation products, which
are expected to be lower cost, provide for smoother network operations
and to offer more management features.
REFFERENCE
I got The above Information And Refference From Under
the guidance of several Textbook & Authors name as :
See the following documents for additional information on ATM
technical descriptions, standards, and benchmarks:
IEEE 802-1990 IEEE Standards for Local and Metropolitan Area
Networks: Overview and Architecture.
Telcordia GR-1110-CORE.
Performance Evaluation of Connection Admission Control
Techniques in ATM networks, B. Jamoussi. S. Rabie, and O. Aboul-
Magd, GLOBECOMM96, November, 1996.
And I get all information from the web 'Nortel Confidential'.
Atlast I thanks our H.O.D sir Mr. Balaji padhi For their
advice to refer to this Textbooks.
