Avionics Bus Technology: Which Bus Should I Get

On?
Tony Ricker
Engineering Department
Avionics Interface Technologies
Beavercreek, OH
tonyr@aviftech.com

both the hardware and data formats required for bus

I. INTRODUCTION: transmission. ARINC 429 is most widely used in
commercial aircraft manufactured by Boeing, Airbus,
Today's commercial and military aircraft utilize many Embraer, Bombardier, and many others. ARINC 429
different avionics data and video bus technologies use is ubiquitous, and it is still designed into new
including MIL-STD-1553, ARINC 429, ARINC 664, programs that do not have the requirement to update
Fibre Channel, Time-Triggered Ethernet, and ARINC to a different network.
818. Avionics designers want to insert high-speed
B. Overview of ARINC 429 Protocol
networking technology while still utilizing data bus
technologies already integrated and proven. The ARINC 429 is a data link supporting 100 kilobits per
array of data buses can be confusing, and it isn’t clear second (kb/s), 25 kb/s, and 12.5 kb/s options.
which bus is best for different types of applications. ARINC 429 data is transferred in 32-bit words. The
This paper presents an overview of the most common data contained in each 32-bit word is identified by an
avionics network and data bus technologies, 8-bit label ID. Other optional data fields are provided
including the strengths and weaknesses of each. It to further identify the payload data including data,
explores key characteristics that determine which bus sign/status matrix (SSM), and source/destination
is best given a specific technical requirement. For identifiers (SDI) fields. ARINC 429 uses odd parity.
example, ARINC 664 (AFDX) and SAE 6802 (Time- (The last bit of a 32-bit bus word is a parity bit.)
Triggered Ethernet) are both deterministic versions of However, some avionics applications have chosen
Ethernet, so how does one determine which is more even parity, and in some cases the parity bit may be
appropriate? Several buses can be used for video used as an additional bit for data.
applications, but which bus provides the most
comprehensive set of features for demanding video
links? This paper will compare and contrast the
different data buses to help the readers determine the
best bus for their applications given their technical
requirements. Figure 1: ARINC 429 32-bit bus word
Table 1 gives a brief overview of each protocol, and
the following section give a detailed description of The hardware consists of a single transmitter
each protocol. connected to one or multiple receivers. Up to 20
receivers can be connected to a single transmitter.
All ARINC 429 data links are unidirectional with no
acknowledgment of receipt of data, which means no
II. ARINC 429 handshaking is involved. A differential (2-wires)
bipolar return-to-zero encoding scheme is used.
A. Introduction
Unlike modern networking protocols, this is a
ARINC 429 is a specification that defines the signaling standard; thus the sender always sends to
physical, electrical, and functional characteristics for the line, and the recipients always read from it. If no
the transfer of digital data between avionics systems data is available for sending, the line is set to zero
on commercial aircraft. ARINC 429 is also known as voltage.
the Mark 33 DITS Specification. The protocol, which ARINC 429 has the concepts of transmitter channels
was released in 1977, is a simple serial data link and receiver channels. The transmit channel is
between avionics equipment and systems and defines
Avionics Bus Technology: Which Bus Should I Get On?
©2017 Avionics Interface Technologies

978-1-5386-0365-9/17/$31.00 ©2017 IEEE

responsible for sending the labels and data and the ARINC 429 is so widely adopted, COTS test
specific data rates for the line-replaceable unit equipment has become increasingly cheaper.
(LRU). The receiver channel is responsible for Compared to other protocols, ARINC 429 is still a
receiving and processing the data. As mentioned low-cost solution sending for limited amounts of low-
above, all data is centered around the concept of the speed data repetitively. But it has limiting factors
label ID. Due to the 8-bit label field, there is a limit that is making Ethernet-based protocols an effective
of 256 defined labels. Label guidelines are provided replacement.
as part of the ARINC 429 specification for different
equipment types. Each aircraft will contain a number
of different systems, such as flight-management
computers, inertial reference systems, air data
computers, radar altimeters, radios, and GPS sensors. III. MIL-STD-1553
For each specific type of equipment, a set of standard
parameters is defined, which is common across all
manufacturers and models. A. Introduction
The simple architecture, almost point-to-point wiring, MIL-STD-1553 is a military specification published
provides a highly reliable transfer of data. Typically by the U.S. Department of Defense that defines the
star or bus-drop topologies are used, but multiple mechanical, electrical, and functional characteristics
design configurations are used as well (Figure 2). of a digital time division command/response
multiplexed serial databus. The standard, which was
released in 1973, was first published as a U.S. Air
Force standard for the Lockheed Martin’s F-16
program. Currently, MIL-STD-1553 is ubiquitously
used for a large number of military and civilian
avionics applications. Due to its low latency
command/response protocol, MIL-STD-1553
continues to be designed into new programs such as
the Airbus A350, F-35, and several UAV programs.

B. Overview of MIL-STD-1553 Protocol

MIL-STD-1553 defines redundant shared-data-bus
architecture with a single bus master (bus controller)
and supports up to 31 terminal devices (remote
Figure 2: ARINC 429 system topologies terminals). Additionally, an unlimited number of
monitoring devices (bus monitors) are allowed on the
bus.

C. Conclusion
ARINC 429 is widely used in avionics for
commercial aircraft. For example, it is used by
Embraer (E-Jet, ERJ, and Legacy), Boeing (737, 747,
757, 767, 787), and Airbus (A330, A340, A380, and
A350), to name a few. Compared to other
commercial protocols like ARINC 664, its simplistic
architecture allows for high reliability and easy
testing but can also pose a challenge when designing
complicated systems with dense interlinking, which
amounts to a lot of weight in cabling. The high- and
low-speed data rates, 100 kHz and 12.5 kHz, are also
limiting factors for ARINC 429 and explain why Figure 3: A MIL-STD-1553 bus architecture
newer data buses and networks often take its place.
However, ARINC 429 has been around for many
years and adopted for many platforms. The number C. Bus Controller
of existing technologies and LRUs that currently use The bus controller’s main function is to designate the
ARINC 429 is extremely wide ranging, and since master bus responsible for providing all the data flow
Avionics Bus Technology: Which Bus Should I Get On?
©2017 Avionics Interface Technologies
control for all transmissions on the bus. All
information is communicated in a command/response
mode, where the bus controller initiates all
communication by sending command words while
the remote terminals reply with status words. The
bus may support multiple bus controllers, but only
one can be active at a given time.

D. Remote Terminals
The remote terminal (RT) is a device to interface the
subsystems with the 1553 data bus. The RT is
responsible for receiving commands from the BC and
reacting accordingly. The RT cannot transmit data on
the bus unless the BC commands it to do so. There
can be up to 31 remote terminals on the data bus, and
each RT can have up to 31 subaddresses.

E. Bus Monitors
Bus monitors are passive devices on the bus that are
allowed to monitor and capture all the bus traffic. Figure 4: MIL-STD-1553 word formats
Bus monitors never transmit any information on the
bus. They only function to store all or selected MIL-STD-1553 is a highly reliable, serial military
portions of the data transmitted between the BC and avionics field bus with an extremely low error rate of
RTs as well as to collect electrical or protocol errors. 1 word fault per 10 million words. It utilizes a
There are three types of commands that can be shielded twisted pair transmission medium with a
initiated by the bus controller: Manchester encoded signal. The data rate for the
• ƵƐ ĐŽŶƚƌŽůůĞƌ ƚŽ Zd ƚƌĂŶƐĨĞƌ ĐŽŵŵĂŶĚ MIL-STD-1553 bus is 1 megabit per second (Mb/s).
;ƐĞŶĚƐĚĂƚĂƚŽZdͿ
• ZĞŵŽƚĞ ƚĞƌŵŝŶĂů ƚŽ ďƵƐ ĐŽŶƚƌŽůůĞƌ F. Conclusion
MIL-STD-1553 is a very well established, well
ƚƌĂŶƐĨĞƌĐŽŵŵĂŶĚ;ĐŽŵŵĂŶĚƐZdƚŽ
proven, serial data bus system for military real-time
ƐĞŶĚĚĂƚĂƚŽͿ system applications and has qualities well suited to
• ZĞŵŽƚĞ ƚĞƌŵŝŶĂů ƚŽ ƌĞŵŽƚĞ ƚĞƌŵŝŶĂů command-and-control applications in harsh
environments. MIL-STD-1553 is predominantly
ƚƌĂŶƐĨĞƌĐŽŵŵĂŶĚ;ĐŽŵŵĂŶĚƐĂŶZd used for military applications. The applications
ƚŽƐĞŶĚĚĂƚĂƚŽĂŶŽƚŚĞƌZdͿ typically are for avionics data buses and weapons
data buses. MIL-STD-1553 is generally used on
MIL-STD-1553 utilizes a redundant bus architecture. platforms including submarines, tanks, target drones,
The bus controller tries commands to the remote missile and satellite systems, land-based and launch
terminals on one of the redundant buses. If no vehicles, and the International Space Station and
response is received, the bus controller does a retry of other space programs.
the command on the other redundant bus. The MIL-STD-1553 has been around for more than 40
redundancy is not automatic, but is controlled by the years and is fielded on a number of different
application. A complete message from the BC to the platforms. Because it has been so widely adopted,
RT involves a command word, data word(s), and a like ARINC 429, the amount of existing technology
status word. A maximum of 32 data words can be and LRUs that currently use MIL-STD-1553 is
sent or received with a given transfer. extremely wide ranging, and for this reason COTS
test equipment has become much cheaper. It
definitely has many advantages from maturity,
redundancy, and reliability but also has some
shortfalls with the slower speed, limitations on data
size with 32 words per transfer, and low flexibility
compared against networks. Additionally the

Avionics Bus Technology: Which Bus Should I Get On?

©2017 Avionics Interface Technologies
electrical characteristics of the physical medium is
tied closely with the protocol—unlike, say, Fibre
Channel, which allows for easier adaptation of
different upper layer protocols on the same copper or
fiber medium. Fibre Channel aimed to separate the
physical medium from the protocol for greater
flexibility. (This will be discussed later in the Fibre
Channel section.) MIL-STD-1553 remains an
effective and reliable solution and still used on the
newest aircraft today, but due to some of the
limitations discussed above, Fibre Channel and
Ethernet protocols are been used more and more for
upgrades and new designs.

IV. ARINC 664 / AFDX®

A. Introduction
As discussed above, many commercial aircraft
initially used the ARINC 429 standard for avionics
communication. However, ARINC 429 has Figure 4: Avionics Switched Network Topology
limitations as a unidirectional bus with data rates of
only 12.5 kHz and 100 kHz and a 32-bit data word. The end system is the device whose applications
ARINC 664/AFDX® was designed as the next- access the avionics Switched Ethernet to send and
generation aircraft data network. A commercial receive data via the network. Each end system has a
standard (ARINC 664) and an aircraft vendor- direct, bi-directional connection to a switch. There is
specific implementation known as Avionics Full an optional second bi-directional connection to
Duplex Switched Ethernet (AFDX®) have been another switch that is used for the redundant
developed that defines the topology and use of communication path. The switching technology
Switched Ethernet in an avionics application. ARINC ensures that the connection and bandwidth required
664 defines the use of IEEE 802.3 Ethernet and to move data from one end system to another is
Internet Protocols (IP, UDP, SNMP, etc.) for avionics available. The switches are responsible for managing
applications. AFDX® is a specific, deterministic (or filtering) data traffic. The switch performs
implementation of an ARINC 664 network that has connections based on the virtual link ID located in
been developed by Airbus for the A380, A350, and the MAC protocol layer of the data frame and
A400M aircraft programs. Other ARINC 664, implements a special function to ensure that each
deterministic implementations are used on the Boeing transmitting port sends its defined virtual link frames
787, COMAC ARJ21, and Bombardier C Series within its allotted window. As AFDX is built on
aircraft programs. standard Ethernet, the frame structure resembles that
of an Ethernet frame.
B. Overview of ARINC 664 Protocol
As mentioned above, a commercial standard called
ARINC 664 has been developed to define Switched
Ethernet for commercial aircraft applications. It is
baselined on standard Ethernet to take advantage of
reducing costs and development time, but differs
from standard Ethernet in two key factors that are
essential for an avionics communication network:
redundancy and determinism.
Avionics Switched Ethernet is a closed network
topology. The elements of the topology are end
systems, switches, or the connections.

Avionics Bus Technology: Which Bus Should I Get On?

©2017 Avionics Interface Technologies
 time period consisting of the BAG plus jitter. Jitter
has a direct relationship with the network usage,
Figure 5: AFDX® / ARINC 664 p7 Frame Structure meaning more VLs sharing a physical link equals
higher maximum jitter. If there is too much jitter
associated with a frame transmission, the frames sent
outside the window tolerances are discarded.
ARINC 664 uses the concept of a virtual link (VL) to
define logical channels through a Switched Ethernet
network from a single transmitter to one or more
receiving end systems. The virtual link is the basis of
the avionics Switched Ethernet protocol. Each end
system will exchange data frames via a defined
virtual link. All switching for a frame from an end
system transmitter to a receiver end system is based
on a virtual link. The virtual link defines the Figure 7: Bandwidth allocation gap
unidirectional connection from one source end
system to one or more destination end system(s).

C. Conclusion
ARINC 664 has been installed on newer aircraft
platforms. Boeing and Airbus have developed
aircraft utilizing this Switched Ethernet architecture.
Being based on common Ethernet, there are many
advantages including cost, weight, development time,
and higher speeds (10/100/1000 Mb/s) over some of
the other protocols. ARINC 664 includes the
properties of redundancy and determinism required in
aircraft network systems. However, ARINC 664 is
not deterministic to less than 1 millisecond, which
has kept the protocol from being used on military
applications. Additionally, it is licensed and owned
by Airbus, which makes the adaptation of the
protocol more difficult. It is a new avionics network
Figure 6: Virtual Links
technology that is being adopted on more and more
commercial aircraft designs.
The VL is identified in a 16-bit field of the
destination Ethernet address of frames. In addition to V. SAE 6802 (TIME-TRIGGERED ETHERNET)
providing a logical path through the network, VLs
also provide the mechanisms to allow the Ethernet A. Introduction
network to be considered deterministic. Each VL is A new SAE standard called AS6802, also known as
characterized by a bandwidth allocation gap (BAG)
Time-Triggered Ethernet, expands on classic IEEE
and a maximum allowed Ethernet frame size. The 802.3 Ethernet and provides a foundation for the
BAG defines the minimum time distance between integration of services to meet avionics time-critical
two consecutive frames on the VL, and the requirements. Time-Triggered Ethernet is ideal for
transmitter is required to adhere to this speed limit. technology insertion because it is designed for
The network switches have the responsibility to integration with existing networks. The AS6802
monitor and police the VLs to ensure that the
standard combines a proven deterministic, fault-
transmitters are not violating the BAG. As a result, tolerant and real-time technology with the flexibility,
the VL concept allows system designers to a partition dynamics, and a legacy of “best-effort” Ethernet and
and prioritize the resources of the shared Ethernet is therefore well suited for all types of applications
network to provide a guaranteed minimum level of and domains. Initial supporters of SAE AS6802
service. The designer must also take into account the
standardization project include Lockheed Martin,
jitter, since the window in which VLs are sent is a

Avionics Bus Technology: Which Bus Should I Get On?

©2017 Avionics Interface Technologies
Bombardier, Embraer, General Dynamics, ϯ͘ ,ŝŐŚ ƌĞůŝĂďŝůŝƚLJ ĂŶĚ ĂǀĂŝůĂďŝůŝƚLJ ŽĨ ƚŚĞ
Honeywell, and TTTech.
ŶĞƚǁŽƌŬ͕ĂŶĚ
ϰ͘ ^ĐĂůĂďůĞĞdžƉĂŶƐŝŽŶŽĨƚŚĞŶĞƚǁŽƌŬ
B. Overview of Time-Triggered Ethernet Protocol
The standard defines basic time synchronization AS6802 is defined to be completely compatible with
services that are transparently integrated with IEEE 802.3 Ethernet standards. This allows easy
standard IEEE 802.3 Ethernet’s message-based integration with existing networks of conventional
communication infrastructures. PCs, web devices, multimedia systems, real-time
On the AS6802 network, three different traffic types systems, and safety-critical systems on the same
(quality of service) messages are defined: backbone network.
• ĞƚĞƌŵŝŶŝƐƚŝĐƚŝŵĞͲƚƌŝŐŐĞƌĞĚ;ddͿƚƌĂĨĨŝĐ͕ Critical applications have a guaranteed fail-safe with
redundancy enabling fault tolerance. Critical control
• ǀĞŶƚͲĚƌŝǀĞŶ Žƌ ƌĂƚĞͲĐŽŶƐƚƌĂŝŶĞĚ ;ZͿ
systems, audio/video and standard LAN applications
ƚƌĂĨĨŝĐ͕ĂŶĚ can safely coexist in one deterministic Ethernet
• ĞƐƚͲĞĨĨŽƌƚ ;Ϳ ƐƚĂŶĚĂƌĚ ƚŚĞƌŶĞƚ network. The system remains fully functional even if
a failure occurs – supporting single or double faults.
ƚƌĂĨĨŝĐ Deterministic Ethernet uses redundant network paths
and guardians to disconnect with faulty segments or
ports. The redundancy management defined in the
The traffic type of the message is identified based on AS6802 specification is a key difference between
a message’s Ethernet MAC Destination address. deterministic Ethernet and other “safe” Ethernet
Messages from higher layer protocols, like IP or networks.
UDP, can be “made” deterministic without An AS6802 network can be configured to operate as
modifications of the messages’ contents. a fully redundant, multi-domain application. This
Events in AS6802 systems occur at predefined times redundancy enables messages to be sent
with single-microsecond precision. This includes the simultaneously over two separate physical networks.
transmission of TT messages. The AS6802 network Upon receipt of duplicate messages, the receiver
system design predetermines when the TT messages simply passes the first “good” message along to the
are transmitted by which participants and who shall application, while the duplicate message is thrown
receive them. This ensures that the network processes away. This method of reception is often referred to as
TT messages without collisions – that is, without data “first valid frame wins.” This provides the highest
delays caused by collisions. This makes AS6802 degree of fault-tolerance for safety critical
suited for applications of the highest safety critical applications.
level. In addition to fault tolerance, self-stabilization
AS6802 network switches allow the simultaneous properties are built in. For example, the
distribution of TT messages to groups of end systems synchronization will be re-established even after
or the connection of unsynchronized Ethernet repeated failures in a multitude of devices in the
networks. AS6802 networks can be divided into distributed computer system.
domains of smaller application-specific sub-
networks, and the communication between these
domains can be managed. C. Conclusion
The AS6802 definition does not affect the physical The AS6802 standard defines deterministic
Ethernet layers of the protocol. The network can be communication over Ethernet networks. AS6802
used on copper- or fiber-based networks. In addition, provides all necessary mechanisms for applications
it works on 10/100 Mb/s networks as well as on 1 as diverse as classical web services to time-critical,
Gb/s networks. safety-critical and security-critical networks.
The AS6802 standard includes four major design Although defined as a standard, Time-Triggered
objectives: Ethernet is a proprietary protocol, which makes it
ϭ͘ ^ĞĂŵůĞƐƐĐŽŵŵƵŶŝĐĂƚŝŽŶŽĨĂůůƚŚĞƌŶĞƚ more difficult to adopt for open systems. It can be
expensive and does not have the proven maturity and
ŶĞƚǁŽƌŬ ĂƉƉůŝĐĂƚŝŽŶƐ͕ ŝŶĐůƵĚŝŶŐ
reliability of MIL-STD-1553, ARINC 429, ARINC
ĚĞƚĞƌŵŝŶŝƐƚŝĐĂƉƉůŝĐĂƚŝŽŶƐ 664p7, or Fibre Channel. However the AS6802
Ϯ͘ ^ƵƉƉŽƌƚĨŽƌƐĂĨĞƚLJĐƌŝƚŝĐĂůĂƉƉůŝĐĂƚŝŽŶƐ network has the ability to meet the needs of
extremely demanding industries.

Avionics Bus Technology: Which Bus Should I Get On?

©2017 Avionics Interface Technologies
VI. FIBRE CHANNEL

A. Introduction
Fibre Channel is being implemented as an avionics
communication architecture for a variety of new
military aircraft and technology upgrades to existing
aircraft. The Fibre Channel standard defines various
network topologies and multiple data protocols.
Some of the topologies and protocols (ASM, 1553,
RDMA) are suited for avionics applications, where
the movement of data between devices must take
place in a deterministic fashion and needs to be
delivered very reliably.

Figure 8: Fibre Channel overview

B. Overview of Fibre Channel Protocol
A very high view of Fibre Channel shows it is an Fibre Channel is described in the standards as a stack
accepted international standard. Within the of architectural levels. For avionics platforms, we
American National Standards Institute (ANSI) is a will focus on four of the seven architectural levels
group called the InterNational Committee for pictured below.
Information Technology Standards (INCITS), which
is home to a task group called T11. Defining Fibre
Channel is the responsibility of the T11 task group
committee. Fibre Channel is a commercial-off-the-
shelf (COTS) technology enjoying all the benefits of
a $3 billion to $5 billion commercial marketplace
defined as the Storage Area Network (SAN) market.
Fibre Channel has been designed to be a
communication protocol between host processors and
secondary storage elements like disk drives and tape
drives. Even in the commercial marketplace,
deployed Fibre Channel systems more closely match
mil-avionics systems because the commercial world
greatly values the integrity of their core corporate
data files and reliable access with ultra-high
availability.
Other standards tie the physical medium closely with
the protocol. Fibre Channel has been designed to be
a universal carrier of information and crafted to
easily map to other protocols. In fact, Fibre Channel
does not have a native “command set” of its own. Figure 9: Fibre Channel Architectural Levels
For an application to utilize the Fibre Channel
physical and logical transport layers, someone must FC-0, the bottom level, describes the physical
map an existing command set to it, like SCSI or interface for Fibre Channel. This level specifically
1553, or invent one of their own like the Anonymous defines the speeds, transceivers, connectors, and
Subscriber Messaging Protocol (ASM) utilized on cabling. Currently there are three baud rates shipping
some advanced avionics programs. in commercial and military avionics Fibre Channel
products. These are 1.0625 gigabaud, 2.125
gigabaud, and 4.25 gigabaud.

Although the technology is called “Fibre Channel,”

the physical media variants include copper along with
long-wave and short-wave optics utilizing both
single-mode and multimode cabling. Fibre Channel

Avionics Bus Technology: Which Bus Should I Get On?

©2017 Avionics Interface Technologies
systems are flexible enough that every physical link Switched-fabric topology allows up to 14 million
may be chosen based on that link’s requirements and ports to be connected together in a generic
not on the presence of other media types elsewhere in centralized switch environment.
the system.
The FC-2 level is interesting because it describes
how Fibre Channel manages the flow of information
between two ports. This level was designed by the
standards body to allow ease of mapping to Upper
Level Protocols (ULP) through an intermediate
mapping level (FC-4).
Fibre Channel supports three pure topologies and one
hybrid. Fibre Channel supports the point-to-point,
arbitrated loop, and fabric switched topologies. The
hybrid topology supported is arbitrated loops
attached to switch ports. Take note, in all topologies
a Fibre Channel port’s transmitter is only ever
physically attached to one other port’s receiver. This A simple solution for higher availability and greater
simplification of the link allows for highly reliable fault tolerance utilizing any of these topologies
transmissions, bringing system errors in even ultra- employs redundant physical ports on a node attached
high-speed links to acceptably low rates. to redundant topologies.
For instance, Figure 10 below depicts a single node
C. Point-to-Point with four Fibre Channel physical port connections.
Two can connect to two redundant switches and two
Point-to-point topology employs two ports connected
can be tied into two redundant arbitrated loops. This
together without any switches present.
would be a very fault tolerant system.

Arbitrated Loop
Arbitrated-loop topology provides a low-cost
attachment of one to 126 ports in a loop fashion.
Two ports is the practical minimum, and the switch is
distributed into each port. Arbitration is done among
the nodes to determine who can transfer data.
Figure 10: Fault Tolerant Node

This also illustrates that a single physical Fibre

Channel link is comprised of two independent fibers:
a dedicated transmit and a dedicated receive. This
means that the port is full-duplex capable of
receiving and transmitting concurrently. In fact, in a
switched topology, a port can be receiving data from
a source port while sending data to a completely
different destination port.

Switched Fabric

Avionics Bus Technology: Which Bus Should I Get On?

©2017 Avionics Interface Technologies
 D. Frame Structure is sent across the media. Typical QoS features
The Fibre Channel frame is smallest unit of include guaranteed bandwidth, guaranteed latency,
information transferred between two ports. Frame acknowledged delivery, notification of non-delivery,
header fields carry additional control information end-to-end flow control, and guaranteed in-order
about start, middle, and end of sequences and delivery of frames within a sequence.
exchanges, sequence initiative, and connection Class 3 is the dominant user class of service deployed
boundaries (Classes 1 and 4). Figure 11 illustrates in both the commercial and mil-aero markets. It is a
the structure of the Fibre Channel frame, detailing the best effort packet-switched service that resembles a
frame header. datagram service with no attendant QoS features.
Class 2 has been implemented by many vendors of
Fibre Channel hardware with a view to the future. It
also is a packet-switched service but it has end-to-end
flow control with acknowledged delivery and
notification of non-delivery. It does not guarantee
bandwidth or latency of messages.

F. Conclusion
With the advantages that Fibre Channel offers, it is
currently being implemented on new military
platforms and used as a technology update on others.
Examples include the F35/JSF, F18, F16, E2C, and
Figure 11: Fibre Channel Frame Structure B1B. With the way Fibre Channel was architected
and it being easily adaptable to multiple protocols
such as ASM, RDMA, MIL-STD-1760E, etc., it is a
very attractive solution for avionics applications
E. Class of Service where the movement of data between devices must
Another valuable Fibre Channel basic relates to user take place in a deterministic fashion and needs to be
classes of service that Fibre Channel offers. delivered very reliably. Many new upgrades on
Generally, there are four user classes of service military aircraft are being implemented with Fibre
defined in Fibre Channel. However, only two of Channel using MIL-STD-1553 protocol. MIL-STD-
them are widely deployed. Classes of service have to
do with the Quality of Service (QoS) with which data

and Thales each had their own protocols for their

VII. ARINC 818: AVIONICS DIGITAL VIDEO BUS products.
ARINC 818 standardizes the Avionics Digital Video
A. Introduction
Bus (ADVB). It is a protocol for low-latency, high-
In 2005, Airbus, Boeing, and other major aerospace bandwidth digital video transmission in both
companies decided to standardize cockpit display commercial and military applications. The ARINC
interfaces. A new standardization effort began 818 specification was ratified in January 2007 with
through the Digital Video Subcommittee of ARINC. participation from wide range of aerospace suppliers.
The primary driver for the standard was a need to The specification was revised in 2013 to ARINC
consolidate many proprietary standards that existed 818-2 adding the following capabilities:
in the avionics supply chain. For example, display
manufacturers such as Honeywell, Rockwell Collins,

and protocol capabilities found in modern networking

B. Overview of ARINC 818
protocols. ARINC 818 is deterministic with low
The major aim of the ARINC 818 specification was latency and includes error detection. ARINC 818 is
to provide a robust protocol to handle the high built specifically for high-speed video, embedded
bandwidth of modern avionics video systems. Fibre data, and the technical challenges that video
Channel (FC0 and FC1) remains the physical layer encompasses—for example, embedding the timing
for the bus and also offers the advantages of routing needed to drive line-synchronous displays. ARINC
Avionics Bus Technology: Which Bus Should I Get On?
©2017 Avionics Interface Technologies
818 is extremely flexible to accommodate different frame, and asynchronous, which will accommodate
link rates, video resolutions, and pixel-encoded and the most demanding video applications.
ancillary data. It requires only a small ICD (interface
control document). Originally, ARINC 818 was
E. ARINC 818 Packets
developed for high-resolution displays, but it is now
also used for sensors, cameras, video processors, ARINC 818 has precise rules for constructing a
radar, and other video intensive systems. packet. The ARINC 818 standard refers to the basic
transport mechanism (packet) as an ADVB frame. It
is important to refer to these packets as “ADVB
C. High bandwidth frames” rather than simply “frames” to eliminate
At the time ARINC 818 was ratified, the fiber potential confusion with video frames.
channel protocol supported link rates of 1.0625,
2.125, 4.25, and 8.5 Gb/s. Since then, link rates of
14.025, and 28.05 Gb/s have been released, with even
higher speeds planned as the market needs it. For
example, a display at WQXGA resolution (2560 x
1600 pixels @ 24-bit color) at 30 Hz would need a Figure 12: Structure of ADVB Frame
bandwidth of 3,864 Mb/s.
The start of an ADVB frame is signaled by an SOFx
ordered set and terminated with an EOFx ordered set
Table 2. ARINC 818-2 Speeds (Figure 12). Every ADVB frame has a header
Bit Rate (Gbps) Note
comprised of six 32-bit words. These header words
1.0625 FC 1x rate pertain to such things as the ADVB frame origin and
1.5 intended destination and the ADVB frame’s position
1.62 within the sequence. The payload can be video or
2.125 FC 2x rate data. The payload in one ADVB frame can vary in
2.5 size but cannot be greater than 2112 bytes. Finally,
3.1875 FC 3x rate all ADVB frames have a 32-bit CRC calculated for
4.25 FC 4x rate all data between the SOFx and the CRC word for
5.0 built-in error checking. The CRC is the same 32-bit
6.375 FC 6x rate polynomial calculation defined for Fibre Channel.
8.5 FC 8x rate An example of how ARINC 818 transmits color
12.75 FC 12x rate XGA provides a good overview. XGA RGB requires
14.025 FC 16x rate about 141 MB/s of data transfer (1024 pixels x 3
21.0375 FC 24x rate bytes per pixel x 768 lines x 60 Hz). Adding the
28.05 FC 32x rate protocol overhead and blanking time, a standard link
rate of 2.125 Gb/s is required. ARINC 818 packetizes
The 6x, 12x, and 24x speeds were added to video images into ADVB frames. An ADVB frame
accommodate the use of high-speed, bi-directional is defined in Figure 13, where the maximum size of
coax with power as a physical medium. The 5 Gb/s the payload is 2112 bytes. Each ADVB frame begins
rate was added to supported by certain FPGAs with a 4-byte ordered set, called an SOFx (start of
implementations. frame), and ends with an EOFx (end of frame).
In addition to the above speeds, an interface control Additionally, a 4-byte CRC is included for data
document (ICD) can specify other rates for a specific integrity.
data-only return path implementation. For example,
a camera might have a low-speed control link that
does not need even the FC 1x rate.

D. Optimized for Video

ARINC 818 is designed as a video and data protocol
and includes many features to accommodate almost
any video format imaginable, including RGB, mono,
YCbCr, and other pixel-packing methods. Four
synchronization classes are defined: pixel, line,

Avionics Bus Technology: Which Bus Should I Get On?

©2017 Avionics Interface Technologies
 or port aggregation using Ethernet. In most
implementations, the input is split at the transmitter
device into two or more ARINC 818 links and then
reassembled at the receiver for display or recording.

H. Networking
Because ARINC 818 uses Fibre Channel as the
physical layer and the protocol has support for source
and destination ID in the headers, networking is
straightforward. Repeaters, routing, and fanout
topologies are all possible. This allows a great deal
of flexibility in the design of an overall avionics
display system. However, ARINC 818 will not work
on many Fibre Channel devices due to the bi-
directional requirements of the base protocols.
Figure 13: Color XGA example sequence of ADVB ARINC 818 switches are on the market today.
frames

I. Copper Physical
Each XGA video line requires 3072 bytes, which
exceeds the maximum ADVB payload length; The vast majority of ARINC 818 implementations
therefore, each video line is divided into two ADVB use optical fiber. Important additions to the ARINC
frames. Transporting an XGA image requires 1536 818-2 standard (over ARINC 818-1) paved the way
ADVB frames per image. The first ADVB frame, for improved copper physical layers. Specifically
which carries the Object 0 data and the container envisioned was the use of newer active equalizer
header, brings the total to 1537 ADVB frames. Video chips to greatly improve bandwidth and distance on
and line timing is adjusted by inserting idle coaxial cable. Included in ARINC 818 are methods
characters. to implement a return communication path (from
video receiver to video source) on the same coaxial
cable. Therefore, a complete camera or sensor
F. Low Latency interface can be achieved with a single high-
One of the most important features of ARINC 818 is bandwidth coaxial cable.
the ability to deliver uncompressed video with very
low latency. In many implementations the latency is
J. ARINC 818 Conclusion
just a few video lines. Low latency is important in
real-time cockpit displays and especially in Heads- ARINC 818 continues to be adopted across a broad
Up Displays (HUD) where differences in the HUD spectrum of aerospace programs due to its robust
display images and real-world background can cause error checking and low latency and to high
vertigo or motion sickness in the pilot. bandwidths required for displays, cameras, and
Latency is generally determined by the sensors. It is being used around the world for both
implementation. In some cases, the image is civilian and military aircraft in new development and
streamed through FIFOs and can be almost real-time. upgrade programs. As demonstrated by the active
Other implementations use two image buffers and participation in the development of ARINC 818-2,
display one while the other is filling (“ping and ADVB has wide industry support from aircraft
pong”) giving a latency of a single frame. At 30 Hz, manufacturers and suppliers. With the addition of
this equals latency of 33 milliseconds. At 60Hz it is higher speeds and support for compression,
16 milliseconds, which is more than enough for even encryption, networking, and sophisticated display
the most demanding applications. In ARINC 818, schemes, ARINC 818 adoption will continue to grow
there are no limitations on the frame rate and even and expand the mission profiles within and beyond
shorter latencies are possible with high frame rates. avionics. However, for data applications, such as
command and control, other technologies may be
superior, such as AFDX or 1553.
G. Channel Bonding
For higher bandwidth applications it is possible to
using multiple channels to carry a video stream.
This is called channel bonding and is similar to link
Avionics Bus Technology: Which Bus Should I Get On?
©2017 Avionics Interface Technologies
 production based in Dayton, Ohio, near Wright
Patterson AFB. AIT is a registered
ISO9001:AS9100C company.
VIII. OVERALL CONCLUSION
All the protocols discussed in this paper have their Great River Technology
place in legacy and new military and commercial Great River Technology (GRT) is the global leader in
platforms. In choosing a protocol, one needs to ARINC 818 and HOTLink II™ systems. Aerospace
identify the important characteristics for the given engineers worldwide tap its expertise and products to
project and choose the appropriate bus accordingly. simplify design, implementation, and testing of
MIL-STD-1553 and ARINC 429 are the most mature mission-critical video and data transmission. GRT
and wide used in military with MIL-STD-1553 and helps them coordinate LADs and conventional
commercial industries with ARINC 429. On the cockpit displays, HUDs, graphics generators, mission
newest aircraft being built today, there are many processors, flight recorders, video switches, infrared
systems that utilize these protocols. However, with and optical sensors, and flight simulators. GRT is
technology evolving and the ability to reach gigabit located in Albuquerque, New Mexico, and is a
speeds rather than megabit or kilobit speeds, ARINC registered ISO9001 company.
664p7 is being used to replace ARINC 429 networks
on commercial aircraft and Fibre Channel is the
protocol replacing MIL-STD-1553 networks on IX. REFERENCES
military platforms. There is a strong desire to move Fibre Channel Framing and Signaling Protocol (FC-
to more Ethernet or Fibre Channel networks and as FS), Rev 1.7, Feb. 8, 2002.
avionics systems progress, this pattern will become Fibre Channel Avionics Environment (FC-AE), Rev
more and more evident. Whereas the majority of 2.6, Feb. 7, 2002.
buses described were specifically developed for data Developing and Testing Networked Avionics
networks, ARINC 818 is optimized for mission- Systems and Devices, ©Avionics Interface
critical video and embedded data. Technologies.
Fibre Channel Testing for Avionics Applications,
AIT ©Avionics Interface Technologies.
Avionics Interface Technologies, a division Switched Ethernet Testing for Avionics Applications,
of Teradyne, Inc. is a leading supplier of avionics ©Avionics Interface Technologies.
data bus modules and a wide array of simulation and “MIL-STD-1553,” Wikipedia,
analyzer products. AIT products include interfaces https://en.wikipedia.org/wiki/MIL-STD-1553.
for MIL-STD-1553, ARINC 429, Fibre Channel, HS- “ARINC 429,” Wikipedia,
1760, ARINC 664 and Deterministic Ethernet. AIT https://en.wikipedia.org/wiki/ARINC_429.
software solutions are used for testing and ARINC ARINC 818-2.
615A and ARINC 615 data Loading. AIT is
headquartered in Omaha, Nebraska, with design and

Avionics Bus Technology: Which Bus Should I Get On?

©2017 Avionics Interface Technologies