Command, control and communications (C³) or, as it is frequently referred to, command, control, communications and intelligence (C³I) is the management infrastructure for defense and war. It is intrinsically a diverse range of activities, incorporating a complex mixture of personnel and man-made systems. At present, we do not properly understand many aspects of either man or his systems, and yet we rely on both to sustain peace and, if necessary, to efficiently conduct war. Current C³I is frequently characterised by high cost, poor functionality, poor decision support, inflexible architectures, and high through-life costs. Although technology is clearly going to provide faster computers with much larger memories, and better displays and MMI, the reduction of software costs and provision for the planned evolution of systems are features that will only come from direct effort within the C³I community itself. The tasks of specifying functions and ensuring that they are achievable and effective requires sound design, analysis and prototyping as the foundation of any major C³I system procurement. Many of the features I have outlined will provide this foundation.

The model development described in this chapter was completed in 1983. Since then the C-process has been successfully employed in a functionalist approach to the design of an experimental command subsystem called the C-station. A simulation of the C-station has been implemented which provides basic C-station features appropriate to antisubmarine warfare (ASW) command. Current developments are aimed at enhancing this basic prototype to incorporate a fuller set of C-station facilities.

This chapter introduces three interlocking concepts which together present a radical option to current air power deployment conventions, which is consistent with threat and technology trends and which at the same time addresses some present day shortcomings. These so-called MOSAIC concepts emphasise survivability and 'movability' as the key issues; integration of offence and defence is seen as a fundamental part of survivability, since neither perfect defence nor perfect offence is a viable military stance.

The design of large-scale C³ systems has in the past relied heavily on expert judgement and intuition with considerable emphasis on improving individual component elements in the existing system. This contribution reviews a flexible and effective method for the study of such systems. The method allows significant achievements to be made in a relatively short time, primarily by maintaining a system-wide over view while taking account of the interrelationships of component elements. In this work a review of the system dynamics modelling technique illustrates the general applicability of the technique to the study of C³ systems. Specific aspects of effectiveness are then discussed, leading to the application of system dynamics the study of information flows and the analysis of transitioning problems. In seeking to improve existing command and control systems or design new ones, it is essential to consider the global implications of any specific changes to the system and to compare the relative impact of alternative designs on the effective ness of the overall system. The system dynamics modelling technique, through the medium of the influence diagram, provides the means of discussing and analysing such problems in a structural and logical way, ensuring as far as possible that the final system is able to perform its intended task effectively through careful planning.

The overall goals of the ISIS project remain well beyond our capability. Neverthe less SDBMS-1 demonstrates that some of these goals are attainable with today's technology, and provides us with an experimental environment in which to study their impact on military planning for command and control. We are convinced that robust, usable spatial information systems for command and control are possible only if their designers and developers are aware of the evolving principles of human-computer interaction. Although improvements in hardware are essential, they must be coupled with a better understanding of the problem-solving skills and requirements of the planner.

In this chapter data management and system design of local area network is discussed.The local area network has brought enormous opportunities to the C³I designer. It allows him to design flexible and modular systems which can be more accommodating of change and enhancement than systems of the past, and which can also exhibit a much greater reliability than was previously possible. By its very nature, a system incorporating a local area network can continue to function even though some of its nodes have failed.

This chapter has described the first iteration of system design and development for a narrowband PR system. The results at all stages have been particularly encouraging. In the second iteration, which is based on a commercial procurement of a 25-station unit of suitable size, weight and power consumption, further refinement and field evaluation of the distributed algorithms will be made.

The concepts of C³ in tactical warfare have existed for many centuries and the need for effective communications to achieve responsive command and control is well appreciated.

With the introduction of computer-based systems in the field of command, control and communications, a vastly increased flow of information is observed within and between the services, both nationally and internationally. This stems from the need to co-ordinate activities between different role cells, command centres, countries and so on. The result is an urgent requirement for the exchange of data between systems each of which operates under the control of a different organisation. Such organisations function according to diverse procedures.

A broad range of civilian OSI application standards is emerging which are potential candidates for use in military C³I systems. The long term benefits of being able to use civilian standards in military C³I systems are such as to make it sensible to continue the process. This chapter develops some of themes in the context of the upper-layer standards. It also discusses the structure of the standards making up these layers of the architecture and their relevance to military C³I systems.

Many military computer systems hold classified information and must be protected from attack. When connected to communications networks these computers will be exposed to potential threats from a very wide population, especially when the networks are interconnected to public switched networks. The technology to communicate with and gain access to these computers is now available in high street shops, and there is a growing band of expert 'hackers' who possess and share expert knowledge of computer systems in order to break in to computer systems for fun. Part of this work has been an extension of the OSI reference model to incorporate security features as a precursor to enhancing OSI protocols. The ultimate aim is to ensure that open systems can be made secure without deviating from published ISO standards.It has been shown that the ISO OSI security addendum to the reference model addresses all aspects of current military security policies where they impact on computer communications. It also addresses issues of concern which are not currently covered by national or NATO security policies.

All three naval engineering standards discussed here have been formally issued. The use of the OSI model as a discipline for constructing the standards has been an interesting experience. The OSI definitions have been lacking in some areas, especially in the handling of recipient-selected broadcast messages. Nor do they offer any assistance in supporting time reference between member systems. Some difficulty was also encountered in maintaining a clean distinction between the layers, as will be appreciated by anyone reading the standards. Nevertheless, the use of the reference model was an invaluable aid in the work of the study group.

Technologically, the engineers have recently developed more and more elaborate ways of transmitting, storing, manipulating and presenting information. Information technology has become a separate profession and has developed its own objectives and standards. The human operator appears in command and control systems at the end, the beginning and sometimes at various intermediate stages. The commander at the end is central to the whole business because he is the primary reason for the existence of the system. He uses it to provide information which enables him to make decisions about dispositions and actions. The design of man-machine interfaces is always critical for the reason already mentioned, namely that the interface must facilitate communication between components with fundamentally different characteristics. The human operator side is highly flexible, adaptable by the use of training techniques and easily reprogrammable in terms of revised instructions, but nevertheless there are fundamental limitations in sensory and motor performance and in information handling.

User-centred design is a methodology for creating user-computer systems which give effective performance and are acceptable to users. These two goals are inter related since effective performance helps to make systems more acceptable to users (especially when their lives might depend on it) and acceptability to users helps to make systems more effective. The standard gives advice on systems design directed at both goals. The goals are not treated separately.

In 1979 the US Department of Defense sponsored a series of command and control (C²) colloquia which concluded that there is no adequate foundation for a theory of C² and hence no principles for overall system design and evaluation (see also Chapter 1). Despite the massive investment by NATO in C² systems research development and implementation in the intervening period, a fundamental metho dology for the design of C² has not emerged. This is not surprising, since high-level C² systems are essentially an extension of basic human decision processes by means of procedures, organisations, equipments, situation assessment (threat assessment) and resource allocation. C² systems are among the largest and most complex real time resource management systems known to man; their effectiveness is severely limited by lack of speed, data saturation and the cognitive limits of the human decision maker.

At the beginning of this contribution it was suggested that 'data fusion' was a term which could be misleading when used in relation to C² systems. A number of different processes have been discussed which are necessary in such systems to enable the user to determine the position, identity and behaviour of objects in this area of interest. In short, they are necessary for him to work out what is happening when his information comes from a variety of sources, some of which may be imprecise, intermittent or even contradictory. The differences between such processes as correlation which establishes that two sets of data refer to the same object, and fusion which combines data which have already been correlated, are important because of the different logical and computational approaches necessary. However, the interrelationship of these processes in the overall operation of a C² system is also important in the progression from data collection by sensors through picture compilation and situation assessment to threat evaluation. A study of current systems and those projected for the near future suggests that whereas a large degree of automation has been applied to the correlation process, although not always with complete success, the same cannot be said of the other processes. Identification of non-cooperating platforms remains a key problem area in picture compilation, and the difficult problems of automatic identity processing have been discussed above. Behavioural analysis, situation assessment and threat evaluation are still almost entirely manual processes al though the pace of modern warfare places increasing loads on the operator and, as mentioned above, in some cases it has been necessary to automate, albeit using very simplistic criteria. The operator load in these vital processes makes it essential to provide automated assistance that is accurate, consistent and reliable. Development of such assistance requires a clear understanding of the precise nature of the various re levant processes and the differences between them.

This contribution describes our research programme at ARE to introduce knowledge-based systems to problems in command and control. Although we are most interested in applications for the Royal Navy, the fundamental requirements are common to most command and control systems and we hope the generality of the concepts will be apparent. The purpose of command and control is to achieve some desired objectives using available resources to best effect. Because these resources usually involve many individuals, both men and machines, a considerable amount of communication and information handling, including intelligence, must take place hence C³I. The reason for the complexity of C³I systems is that they have to tackle very large problems in a complex unpredictable environment. It is these characteristics which generate the need for so much human skill, judgment and intuition and which have caused computers to make such little inroad into the more intellectual processes in command and control. With the arrival of expert systems, however, there is renewed enthusiasm to encapsulate human expertise in computer software, and C³I is an obvious field for attention.

This chapter has described the approach taken in building an expert system for air defence threat assessment. The nature of this problem has required the development of a new programming language called BLOBS. A prototype version of the threat assessment expert system has been built using this framework with some success but contains only a subset of the rules that will be required in the full system. The nature of the threat assessment task is such that experts have difficulty in verbalising certain parts of their knowledge. This is likely to mean that gaps will exist even in future versions of the system. This contributes to our basic view that initial use of the system is more likely to be as an adviser rather than as an autonomous system.