Chapter 9 Avionics Systems Integration 9.1 Introduction and Background Major avionic systems generally comprise a number of smaller sub-systems which are combined to form an overall system. The combination, interconnection and con- trol of the individual sub-systems so that the overall system can carry out its tasks effectively is referred to as ‘systems integration’. The number of sub-systems which need to be integrated to form a major system can be appreciated from the previous chapter on flight management systems. Itis instructive to review the development of avionic systems and their integration into overall systems in the light of the technology available and the circumstances prevailing at the time. The object is to put the development of today’s advanced systems and the even more advanced systems currently under development in per- spective. In many cases the current concepts and philosophy are not new — often the originators of particular system developments in the past were far sighted in their concepts, but, as always, were limited by the technology available at the time. World War 2 (WW2) resulted in a major growth in the electronic equipment installed in aircraft and the birth of avionics, with the very rapid development of airborne radar systems and associated displays, radar warning systems and ECM, and more advanced autopilot systems exploiting electronics. Installation of the elec- tronic equipment (or “black boxes’), however, was very much on an ad hoc basis due to the very rapid developments and time scale pressures in war time. Some very lim- ited degree of integration between sub-systems was introduced, for example coup- ling the bomb sight to the autopilot — as readers who have seen the film ‘Memphis Belle’ may have noted. In general, however, the systems were “stand alone’ s tems and their integration into an overall system was carried out by specific crew members such as the navigator, bomb aimer or radar operator. The 1950s period saw the emergence of a number of avionic sub-systems (some of which were initiated during WW2) which have since undergone continual devel- opment and now form part of the avionic equipment suite of most civil and military jet aircraft and helicopters. For example, auto-stabilisers (or stability augmentation. RPG. Collinson, Introduction to Avionics Systems, DOI 10.1007/978-94-007-0708-5.9, 459 © Springer Science*Business Media B.V., 2011460 9. Avionics Systems Integration systems), ILS, VOR/DME, TACAN, Doppler, air data computers, attitude heading reference systems, inertial navigation systems, The first major step towards integrating avionic systems was taken in the mid- 1950s with the establishment of the ‘weapon system’ concept. These concepts were incorporated in the 1960s generation of aircraft, some of which are still in service. The concept requires a total system approach to the task of carrying out the mission effectively with a high probability of success. The aircraft, weapons and the avionic ystems required by the crew to carry out the mission effectively must thus be con- dered as an integrated combination. It should be appreciated, of course, that the total system approach is applicable to any project, military or civil. As with many methodologies, however, military applications provided the spur and the initial fund- ing. The very widely used ‘programme evaluation review technique’, or ‘PERT’ networks, and ‘critical path analysis’, for example, were originally developed on the POLARIS missile program ‘As an example of the overall system approach, consider the requirements for a naval strike aircraft. The aircraft must be able to operate from an aircraft carrier in all weathers and be able to find the target and attack it with a suitable weapon (or weapons) with a high probability of success. Operational analysis shows that to minimise the probability of detection and alerting the enemy’s defences, the aircraft needs to approach the target at high subsonic speeds (550-600 knots) at very low level at a height of 100 fi or so above the sea so as to stay below the radar horizon of the target as long as possible. The avionic sub-system spei ons can then be determined from the overall system requirements with an aircraft crew comprising pilot and observer/navigator. Hence in the above example, the avionic equipment fit would comprise © Radar target acquisition in all weather conditions. ‘Doppler — accurate ((4 knots) velocity sensor for DR. navigation. (Note: IN ems capable of accurate initial alignment at sea on a moving carrier were still under development in the early 1960s.) © Attitude heading reference system (or master reference gyro system — UK ter- minology) — attitude and heading information for pilot’s displays, navigation computer, weapon aiming computer, autopilot. © Air data computer ~ height, calibrated airspeed, true airspeed, Mach number information for pilot's displays, weapon aiming, reversionary DR navigation, autopilot © Radio altimeter — very low level weather operation. Navigation computer — essential for mission. Autopilot - essential for reduction of pilot work load. Weapon aiming computer — essential for mission. HUD ~all the advantages of the HUD plus weapon aiming for low level attack; for example, ‘toss’ bombing. Stores management system — control and release of the weapons. © Electronic warfare (EW) systems — radar warning receivers, radar jamming equipment. Essential for survivability in hostile environment. ight profile during attack phase and all9.1. Introduction and Background 461 © Identification system (ident attack by friendly forces. ¢ Radio navigation aids — location of parent ship on return from mis © Communications radio suite ~ essential for communicating to parent ship, co- operating aircraft, etc ition friend or foe — ‘IFF*) — essentia ion. A significant degree of integration was required between the avionic sub-systems. For example, the weapon aiming system required the integration of the HUD, weapon aiming computer, AHRS, air data computer and the radar system. The basic avionic systems specified for a naval strike aircraft for service intro- duction in the 1960s, for example the Royal Navy Buccaneer aircraft, would also be required in a 2010-2020 naval strike aircraft. The systems would need to be of much higher performance, however, and there would 3 in a 2010-2020 aircraft which did not exist in the early 1960s. For example, laser gyro INS, helmet mounted displays, GPS, FLIR, “smart weapons’, etc., and a much greater level of systems integration would be required. The aircraft would also be a ingle crew aircraft. The point being made is the essential continuity of the role of avionic systems and their development, although their implementation changes as new technologies become available. It should be stressed that although the above example is a military one, the same principles apply to civil avionic systems in terms of their role in enabling the mii ion to be carried out safely and effectively with the minimum flight crew. Mos civil avionic sub-systems have also been directly developed from military avionic systems (e.g. air data computer, INS, GPS, etc.) and, until fairly recently, military aircraft generally required a larger number of avionic sub-systems with a higher degree of integration. ‘A major step towards facilitating the integration of avionic sub-systems in ci aircraft was taken in the early 1950s with the adoption of ARINC speci for avionic systems and equipment, ARINC is a non-profit-making organisation in the USA which is run by the civil airlines with industry and establishment rep- resentation, which defines systems and equipment specifications in terms of fune- tional requirements, performance and accuracy, input and output interfaces, envir- onmental requirements and physical dimensions and electrical interfaces. For ex- ample, air data computers, attitude heading reference systems, INS, communication radio equipment, data bus systems, etc. Equipment made to an ARINC specification by one manufacturer should thus be completely interchangeable with equipment made by another manufacturer to the same ARINC specification. The electronic implementation of the two systems can be totally different provided they conform to the ARINC specification in terms of form—fit-function. For example, the LRUs (line replaceable units) must conform to the ARINC specifications dimensionally with the specified rack fixing arrangements, connectors and the pin allocations in the connectors. The systems must also meet the ARINC performance and accuracy pecifications and the environmental requirements in terms of temperature range, acceleration, shock and vibration and EMC. The use of avionic equipment qualified to ARINC specifications thus ensures a competitive situation enabling procurement to be made from manufacturers on a worldwide ba462 9. Avionics Systems Integration The early 1960s saw the development of the first real time airborne digital com- puters, and these were progressively introduced in military aircraft from the late 1960s for tasks such as navigation, mission management, weapon aiming, radar processing and displays processing. These first generation airborne digital computers were very expensive and iti teresting to note that a number of proposals were made during the 1960s to promote the concept of carrying out as many of the avionic sub-system computing tasks as possible with a powerful central digital computer. These proposals were not taken up for a number of reasons, such as: ine 1. Vulnerability ~ ‘all the eggs in one basket’, whereby a failure in the central digital computer affected all the sub-systems sharing its computing facilities. 2. Inflexibility — changes in an individual sub-system could involve changes in the main computer software with possible ramifications and ‘knock on’ effects on all the other sub-systems sharing the computing facility. This was particularly relevant with the computing speeds achievable at the time and the high cost and limited capacity of the non-volatile memory technology which depended on magnetic core stores. (Typical store sizes were 8K to 16K at that time.) 3. Cost and weight of redundant central computer configurations was unaccept- able The availability today, however, of affordable, very powerful processors with large memory capacity and high speed data buses has radically changed the situation. The concept of sharing avionic sub-system processing tasks between a number of pro- cessors with spare capacity to take over particular tasks in the event of a processor failure is now economically attractive. The integrated modular avionics architectures which have been developed for the new generation of both military and civil aircraft which entered service from around 2005 onwards, in fact, exploit these concepts will be discussed later. ‘As stated earlier, many of today’s concepts are not new. It is the technology available at the time which limits their economic exploitation. By the latter half of the 1960s, the development of affordable * s°, that is processors which are sized to carry out one (or possibly two) avionic sub-system computing tasks, had become viable with the development of integrated circuit technology. A number of avionic sub-systems had thus been de- veloped and were entering service in both military and civil aircraft by the end of the 1960s which contained their own internal digital computers; for example, digital waveform generators in HUDs, air data computers and inertial navigation systems Many of the aircraft sub-systems up to the early 1970s, however, were still largely analogue in their implementation with synchro and potentiometer outputs/inputs re- quiring point to point wiring to interconnect them. The interface units carrying out the necessary analogue to digital (A-D) and digital to analogue (D-A) conversions to enable the sub-systems to communicate with each other were inevitably bulky and could exceed the size of the digital computer. The complexity of the inter-unit cabling looms and wiring harnesses can be seen in Figure 9.1 which is typical of a late 1960s/early 1970s aircraft installation, It should be noted that some of thes orientated9.1. Introduction and Background 463 New Wiring Fer Additional Sub-Systems Fig. 9.1 Interconnection of avionic sub-systems before introduction of multiplexed data buses. aircraft are still being operated. The weight of the cable looms could be very signi- ficant and exceed several hundred kilograms. The large number of multi-way plugs and connectors also inevitably degraded the overall system reliability and introduced intermittent faults. The first development to reduce the number of interconnecting wires was the use of time division multiplexing (TDM) of the signals to be transmitted between two units. ‘TDM enables information from several signal channels to be transmitted through one communication system by staggering the different channel samples in time to form a composite pulse train. Each signal channel is transmitted as a serial digital pulse train at a given time slot in a clock cycle. Thus given the clock time informa- tion and the address of the signal (suitably pulse coded), the receiver can then decode and distribute the individual signals. Hence if there are, say, 30 different signals to be transmitted between two units, only two wires are required in principle when the data is multiplexed compared with 60 wires. This type of single-source-single-sink communications link between two units (say A and B) is frequently referred to as an ‘A to B’ link. Early systems, in fact, used three screened twisted pairs of wires to transmit the signal, address and clock data. This was then reduced to two screened twisted pairs of wires. Later single- and multi-source-multi-sink systems, however, operating at higher clock rates encode the signal data and address with the clock data using Manchester II bi-phase encoding (covered in the next section). A single screened twisted pair of wires only is required with such an encoding system to transmit information data rates of up to 1 to 2 Mbits/s. Higher data rates of up to 100 Mbits/s can be transmitted using a suitably screened coaxial electrical cable and information at even higher data rates can be transmitted as coded light pulses using a fibre optic cable. This will be covered in Section 9.2.464 9. Avionics Systems Integration Sub-Systems ‘Sub'Systems Bus 1 2 Controller |---| 9? ff Coupier Unit Fig. 9.2 Interconnections of avionic sub-systems by multiplexed data bus. By the mid-1970s, it became possible to implement many more avionic sub- systems digitally, exploiting task orientated processors and the newly developed micro-processors and so eliminate the analogue computing elements and the ana- logue input/output components such as synchros and potentiometers, etc. It thus became possible to interconnect the individual avionic sub-systems by means digital data bus system. This enabled the systems to communicate with each other and transfer serial digital information using time division multiplexing in conjune- ion with a suitable protocol system to control the data transmission to and from each individual sub-system. This will be covered in the next section but the dra- matic reduction in inter-unit wiring can be seen in Figure 9.2 which illustrates the interconnection of the sub-systems using a MIL STD 1553 data bus system ‘An essential parallel development in the late 1970s and early 1980s was the im- plementation of the complex circuitry required by the MIL STD 1553 data bus sys- tem terminals in LSI (and later by VLSI). Thi tem complexity to be encapsulated in highly reliable IC chips with a ‘chip set’ costing less than some multi-way connectors, The avionic system architectures which had become established by the mid- 1970s are still the standard architectures of current aircraft in service and are re- ferred to as “federated architectures’, Federated architectures essentially consist of a number of interconnected but functionally independent sub-systems with some degree of central computer control of overall operating modes. The advance of di- gital electronics through the late 1970s and 1980s enabled the development of very sophisticated sub-systems which, although still addressing self-contained functional areas, shared information via data links such as the 1553 bus. Federated architectures, built up from individual proprietary designs and av: able from a wide and experienced supplier base, became firmly established as the low risk avionics system design approach. Also, because of the well defined sys-9.1. Introduction and Background 465 tem boundaries and the clear division of areas of respo were universally accepted as the low ri as well as technically. Ina federated architecture, however, there is inevitably some duplication of fune- tions in each of the individual sub- ibility, federated systems ially and contractually stems which makes the overall system heavier and more costly than necessary. LRUs are each designed and optimised separately resulting in many different implementations of similar functions. Items such as pro- cessors, memory, software, interfaces and power supplies are individually optimised for each sub-system and the design and support of these unique proprietary parts drastically increases the overall system development and support costs. The sub- system design approach also limits the introduction of fault tolerance because the proprietary sub-systems cannot easily share resources and the cost of building re- dundancy into every sub-system is unacceptable. The use of proprietary sub-systems also creates additional problems whenever an optimised proprietary sub- i terface must be accessed as part of an upgrade, The maintenance of federated systems is constrained by the proprietary bound- aries of the architecture and requires the complete removal of sub-system black boxes from the aircraft for repair in second, third and fourth line maintenance fa- cilities. Flexible deployment of such systems in military applications was seriously hampered by the need to deploy sophisticated second line, or ‘avionics intermediate shop’ support equipment. With the reducing numbers of new military designs the need for mid-life upgrades focused attention on the high cost of upgrading feder- ated architectures. Essentially, whole systems have to be removed and replaced and a small improvement in system functionality can result in very large costs. Since the late 1980s, avionics R&D programmes in the USA and Europe have been directed towards finding a better way to build avionics. Over this period the concept of modular systems built from a small number of standard module types, housed in standard racking, and communicating over standard data networks has been developed and has become known as modular avionics. However, to achieve the full benefits from modular systems it is necessary to dissolve the sub-s in conventional avionics so that a common pool of spare resources, shared between sub-systems, can be used to improve system availability. Hence, integrated modular avionic architectures offer additional benefits of fault tolerance and application flexibility through reconfiguration. Integrated modular avionic architectures have been accepted as the way forward for both civil and military avionics and systems based on these concepts are now in service, or will be shortly entering service. boundaries which exi: