The EMC implications of using COTS in military systems

The electromagnetic environment in which militaiy COTS electronics will be expected to survive is likely to be more severe than its civil counterpart. Thus using COTS equipment could lead to more vulnerable systems than if only military-specified equipment were used. The problem is getting worse because, as technology advances, devices are likely to become more vulnerable. The paper explains how the difficulty may be contained. It identifies the key differences between the civil and military electromagnetic environments, describes the basic requirements and

standards, and summarises design guidance. Protection and hardening methods are then examined. It concludes that, while the necessary knowledge, techniques and procedures are available, caution is needed in integrating COTS items into defence equipment, not least because savings from using COTS items can be largely lost in the costs of testing and protecting them.

Review o f information

This is a ‘Restricted’ paper by an avionics specialist working for DERA. It concentrates on the potential problems in ensuring EMC for militaiy systems containing COTS items. It provides much useful advice on overcoming incompatibility, but points to the potential cost of ensuring compatibility. The widespread application of its advice by COTS IT hardware producers may, to an extent, be limited by its classification.

Benefits

Identifies key differences between civil and military electromagnetic environments, describes basic requirements and standards, and summarises design guidance. Examines protection and hardening methods.

Problems

Electromagnetic environment in which military COTS electronics are expected to survive is more severe than its civil counterpart giving increased vulnerability.

Other issues

Savings using COTS items can be largely lost in costs of testing and protection.

4.8.5 The application o f COTS technoiogy in future moduiar avionic systems’”''"'

Current avionic systems make extensive use of military specifications and standards but are based on dedicated implementations for individual avionic functions that are expensive to develop, procure and support. Future systems will be based on integrated modular avionic concepts in which all functions are implemented on a common processing platform constructed from standard hardware and software modules. To optimise the application to future modular systems, an architecture based on standardised interfaces will allow modules to use COTS components and standards to achieve affordability and exploit rapid advances in commercial technology. At the same time, approaches are being developed to deal with short commercial time-scales and loss of control over specifications and standards.

The main problems and benefits in terms of system architecture, software, networks and processing must be addressed when designing advanced avionic systems using COTS technologies. An architecture using standardised commercially available interfaces leads to lower costs and leverage from the commercial market-place. However, the avionic community demands solutions tailored to provide scalable, flexible, implementations adaptable to future needs.

The specific technical requirements of integrated modular avionics (IMA) software make it difficult to decide if the benefits of COTS outweigh the drawbacks. Unfortunately, there are no actual examples where through-life cost savings have been demonstrated by using COTS software. Only some inconclusive cost-modelling results exist. The correct approach is to design architectures that cater for COTS software components and to postpone the decision on whether actually to use COTS or custom-developed products until required by a particular aircraft project.

For networks, new COTS technologies often offer significantly better performance at potentially lower costs. However, although COTS can save money on basic design, the savings are less when the upgrade to ‘COTS plus’ is implemented, even in a largish military avionic market. Thus the savings predicted by using COTS networks are much more limited in practice. Moreover, the pace of technical change, coupled with unpredictable market forces, can easily undermine the definition of the military platform's architecture before the in-service date is reached.

The development of military-specific processing technology is prohibitively expensive, and there is no alternative to using COTS solutions. Generally, the potential for directly exploiting commercial processing technologies is excellent, with a wide variety of high performance devices being readily available. However, in some areas, IMA system design is at risk owing to obsolescence and environmental considerations, although these effects can be reduced. When the implementation of application functions is examined in more detail, some features of COTS processing devices are not ideal, and so there are still performance issues. However, with careful design, and accepting that some tailoring of applications will be necessary, it should be possible to cater for these shortcomings.

Review o f information

This paper, co-authored by a group fi’om DERA, deals with the specialist area of avionics but their views may also have application to other types of platform. It highlights the tough environmental specifications faced by avionics equipment and the key issues that are faced in designing such systems using COTS IT, in particular, the architectural approach. It recognises that there is no alternative but to use COTS and suggests that some tailoring of applications will be needed. Some of the cost modelling issues, which arise from using COTS, are identified.

Benefits

Affordability and exploitation of rapid advances in commercial technology are features. Significantly better performance is realisable at potentially lower costs.

Problems

Obsolescence and environmental factors are obvious snags. Unpredictable market forces can easily undermine definition of military platform architecture before the in-service date is reached. Short commercial time-scales and loss of control over specifications and standards are problems.

Other issues