Barry Trimmer will debate the issues raised in this article during his presentation at the EuMW Defence and Security Forum in Manchester on 12 October. To find out more and to register for the Forum visit:

The history of military electronic Intelligence, Surveillance, Target Acquisition and Reconnaissance (ISTAR) since World War II has resulted in large single function systems with increasing performance levels achieved at longer and longer ranges (primarily to protect these expensive assets from evolving threats).1 In almost all countries, successive budget reductions have increased the focus on cost-effective provision of ISTAR where traditional boundaries of militarily specific functions and command chains are called into question. The UK for example has recently imposed some of the most severe reductions consequent on its Strategic Defence and Security Review.

In addition, for airborne ISTAR, the advent of the unmanned aerial vehicle (UAV) has reversed the need for long-range sensor performance to give sensor platform protection.2 These stand-in (partly) expendable systems, together with better, more reliable data links, open new possibilities for cost-effective ISTAR provision. This article will highlight some of the technological implications of these changes and the implications for microwave system requirements.

Drivers for Change

For the ISTAR world, there are two change drivers. One is operational and budgetary "pull." The other is technological, opportunistic "push." So the question becomes – how can I address the changes in "pull" by taking advantage of the opportunities of technological "push?" The major change that will drive the use of new technology is budgetary. To understand the effects this will have, consider the principal sources of cost in the armed services. A whole life cost analysis of large programmes, such as the Watchkeeper Unmanned Air System, reveals that the major costs for the armed forces are in personnel and the associated Defence Lines of Development (DLOD).

Many of these DLODs can be assisted by technology. For instance, the training line can be made significantly more cost effective using forms of virtual reality – particularly where the training is for a semi-virtual operational role (e.g. UAV operator). However, some are fundamentally not amenable to this form of efficiency saving, because they are related to staffing or to equipment types required to satisfy the military need.

The tangible result of this budgetary pressure is a reduction in staffing (examples already being announced in the UK). Generally then, there will be less people in the services to make military functions work. Also, the threats against which national services must be prepared are also changing. The effect is often additive, so that the threat of, for instance, terrorism is added to the conventional threats between countries. It is not clear that this total is very much less onerous than the preceding set.

As a result of the pressures applied to the military budget while maintaining the required capability, a change in approach is demanded. The new environment requires technology to specifically answer the question: how can military capability be maintained with both less people in the services and with less money available for equipment purchase?

Technological Opportunities

A principal result of change in technology over the development of electronic ISTAR has been the convergence of functions leading to lower staffing. For instance, while the earliest radar systems used people to tackle individual functions (such as, height finding and bearing of a target), rapid evolution of technology allowed the automatic combination of these functions. Generally then, the use of technology to achieve function combination is a normal evolution, particularly where no human decision-making is replaced by automation. What I see today is a continuation of this theme, but now the technological convergences will also provide an opportunity for sharing technology across military functions – sometimes between dissimilar platforms in different command chains.

There are two obvious convergences happening in military RF sensing at the moment, between RADAR and Electronic Intelligence (ELINT), and between ELINT and Communications Intelligence (COMINT). I will examine later how these can be used to provide one possible answer.

A second enabling issue is the advent of reliable communications, at least at the higher levels of command. This is also a continuing technological trend where access to a communications layer that connects peer military users makes possible the sharing of functions across formerly separate command chains without requiring significant changes in doctrine. It is reasonable to suppose that this trend will continue to provide normal IT-like connectivity (IP-based communications) right down to the tactical level.

As a result, this is a good time to move away from worrying about the provision of a communications layer, to thinking about what the military user would like to do to share military functions given universal addressing, a reliable carrier and an agreed language. This technology is likely to support the provision of co-operation advantages at higher bandwidths while maintaining the security of military communications in a physically distributed, mobile and at least partially wireless environment. This trend provides an alternative approach to reducing the cost of providing military functions. The two approaches are not necessarily in opposition, but are qualitatively different.

Using New Technology

Technology advances in many diverse areas and with many different methodologies. The key to successfully designing a complex multi-technology system is to understand where the "tipping" point occurs. As an example, the recent proliferation of UAV systems has been enabled by the maturity of automatic flight and mission control, reliable data link technology and persistence of air vehicles. The change wrought by these advances transforms the UAV from an asset of occasional use to one that is constantly in demand by the user. I believe that similar "tipping" points exist in the broader ISTAR world that will allow the user to easily take advantage of multiple technology applications to fulfill their military function.

Common Capability Across Military Functions

I have already referred to the sharing of technology across command chains. It is reasonable to suppose that a doctrine can be evolved that reduces the number of people involved in providing that shared service. For instance, the UK MoD has decided to remove the Nimrod MRA4 from the inventory and is intending to take the ASTOR ground surveillance system out of service in 2015. These represent two unique capabilities to maintain maritime patrol and to provide long-range ground surveillance. It is a good thought experiment to consider how such capability could be created by one system using shared technology or by a combination of systems sharing functions across multiple platforms using the evolved communications backbone.

Lower-cost Platforms

Another thought experiment might be to consider how to use a large number of lower-cost platforms to substitute for a single large platform capability? Clearly, the history of electronic ISTAR has been to continue centralisation of functions in larger platforms that have become so valuable that they cannot be risked. This means that the platform has to be further away from the threat. Typically this has led to, for instance, large surveillance radar systems standing off some distance from the threat and having to manage the increase in power required to combat the well-known 1/R4 performance relationship.

Figure 1 Automated/autonomous co-operation between platforms is enabled by a common communications layer, addressing and language.

Suppose I introduce stand-in platforms, such as UAVs that I am prepared to lose in combat, then the physics of the situation radically changes – potentially not needing the expensive, high power technologies. The combination of UAV technology, reliable data links and developing mission autonomy allows this form of system to be seriously considered (illustrated in Figure 13).

Two questions arise:

  • The question for the technologists is what this distributed system needs in component terms and what are the total cost implications?
  • The question for the user is what geographical coverage shape is really required by ISTAR operations now that coverage can move from the conventional circular shape to an arbitrary shape defined on the disposition of multiple smaller platforms? This would be a key to the cost comparison between approaches.

An interesting alternative example exists in Electronic Warfare where a function, such as Direction Finding (DF), is difficult to engineer in mobile platforms. In fact, the ideal function is geo-location, which is often approached by DF over a manoeuvre that, in turn, can be difficult to achieve. Multiple platforms offer alternative measurement techniques to achieve geo-location, such as Time Difference of Arrival (TDOA), that may be simpler to engineer on each platform.

Figure 2 The simplest possible architecture allows multiple functions to be simultaneously available.

Technology Blocks

I referred to the technical convergences in the RF and microwave domain. I can see the evolution of the RF components to serve a very simple architecture where conversion to or from a digital representation happens very close to the front-end of a system (see Figure 24). There are already early examples of this type of architecture. The progress of conversion technology will simply move us further toward this architecture.

There are two reasons driving the adoption of the simple architecture:

  • The overall cost of (nearly) commodity items substituting for bespoke design.
  • The level of flexibility in the digital domain allowing complex signal processing to compensate for front-end behaviour.

The second reason is worth exploring to illustrate the change in system design and flexibility. Suppose I have an analogue receiver chain that leads to a detection process. If this chain is complex then I have to maintain a high degree of linearity to ensure that the relatively simple detection thresholding sees only the signal of interest and none of the potential artifacts generated by the receiver chain.

The alternative sees conversion to the digital domain happening early in the chain, with artefacts generated in the initial RF conditioning and in the conversion process. Subsequent processes, such as down conversion and filtering, are arbitrarily high quality at relatively low cost.

The key, however, is that the detection process can now be quite complex. For instance, the generation of unintended products is a known pattern that can be allowed for, corrected or filtered. This simple idea is fairly well established in non-coherent EW systems, but the latest coherent system designs point toward allowing this approach to be taken in both radar and EW. Effectively, this makes usable systems out of designs that would have looked impractical in earlier implementations.5


Antennas are the most closely coupled components to the real world and, therefore, are likely to be most affected by the commonality we choose to impose between functions. For instance, if I choose to combine a wideband EW function with a lower bandwidth radar function, then I impose a number of difficult simultaneous requirements that need satisfying. Typically, the radar function requires a fully filled array with spacing that approaches half a wavelength. By contrast, the EW function may not need the same spacing because it does not have to use angle as a fundamental resolution element. The EW function will, by contrast, require consistent behaviour over two or more octaves.

One technical convergence in antennas has to do with the increasing use of phase/amplitude interferometry arrays in EW. These do not need to have quite the same perfect amplitude responses required from conventional amplitude comparison DF, but rather have a combined aperture response which, in essence, is similar to the way a radar system would like antenna elements to behave.

Finally, there is an overarching question for antenna aperture design in a system context. Is the true technical trend defined by creating a single aperture, which is capable of all the required functions, or is it defined by information sharing between dedicated apertures, either on the same platform or between platforms?

Figure 3 Alternative convergence routes depend on a cost, complexity and flexibility balance.

This is an intensely platform and operational-use dependent question.

  • If I think about a combat aircraft then the pressure on space for apertures is so great that a single aperture carrying out several functions is very attractive.
  • By contrast for a surveillance asset, achieving a single aperture is probably not as important as maintaining the flexibility to change functions from mission to mission. For this case, the balance between cost, complexity and flexibility is much less clear (see Figure 3). The jury is out on this question.

RF Conditioning

I am using the term "RF conditioning" to refer to the RF feed into a digital conversion process. I am also assuming that the requirement trends already discussed are likely to include the need for wideband high probability of intercept (POI) EW reception. This is a particularly acute example of the influence of RF components if I consider the convergence between radar-band EW and (relatively) narrow-band radar functions.

The most obvious effect of the trend is the inability to filter out signals in the wide operating band, giving rise to a high dynamic range requirement. In particular, this leads to a need for high levels of device protection, inherently high dynamic range components and predictable behaviour as the RF conditioning is saturated.

Because we cannot provide RF filtering if we wish to maintain high POI, the likelihood of saturation is extremely high. This is particularly true for operation of these converged systems in the presence of very high power emitters, such as those found in naval task groups or in close proximity to civilian communication towers.

The first active component behind the antenna will receive all of this power and must not only survive, but continue to operate in a predictable way so that the presence of other signals (such as the radar return) can be established. This will require either very high speed limiting with the characteristics of "soft" limiting, or naturally higher dynamic range technologies, such as GaN.

The comment about predictable behaviour refers back to the discussion on detection processes. The evolution of systems that utilise complex detection processes and assume knowledge of the distortions of the front-end can become vulnerable to errors in that knowledge, or, for array processing, errors in how that behaviour changes over the array. Repeatable and predictable behaviour as these components enter saturation becomes very important to achieving small signal detection in the crowded modern electromagnetic environment.

Continuing the Evolution

To conclude, I find it interesting to think about the two contrasting trends I have discussed in this article. On the one hand, we have convergence of functions enabled by the increasing ability to digitise close to the front-end of systems, and the evolution of the system architectures of EW and radar functions toward a single implementation.

On the other hand, we have the evolution of co-operation technologies that will allow dissimilar sensors to combine their activities to form a single, but distributed sensor net. This particular trend is likely to be fed by the commercial sector where co-operation technology is a fast-paced development domain. Which of these two trends will dominate or how they will combine in military ISTAR is an unknown, but fascinating topic.


  1. A typical example is the very long-range AEW provision using powerful radar to survey an airspace volume sufficiently large to exclude threats to the radar platform itself.
  2. There are requirements for large area surveillance that may still need long-range detection, but the self-protection requirement is reduced.
  3. The common "natural" language here is the enabler for co-operation, but each sensor/platform must be able to communicate in this natural language – the "I" function. This is becoming a normal, almost emergent capability, thanks to efforts on standardisation.
  4. This simple architecture is the aim point. The degree to which current systems attain this level of simplicity is dictated by their critical requirements. For instance, the very high levels of dynamic range characteristic of the COMINT function will limit the current implementations of ELINT/COMINT convergence.
  5. Taking care not to overload such techniques with too many undesired signals. Like all techniques there is always a limit of applicability.

Barry Trimmer graduated in 1978 from Warwick University, UK, in physics, and in 1979, from Sussex University, UK, with a master's degree in astronomy. He joined the radiation laboratory at EMI Electronics, Hayes, UK, working on radar antennas for the Searchwater radar and naval ESM systems. During the 1980s, he developed RF and system modelling within EMI, leading to system design of ground surveillance, weapon locating and man portable radar systems. In the early 1990s, he participated in the development of combat radar systems and designs for multi-sensor military configurations of civil air platforms. Since 1992, he has led the design of airborne radar, EW, multi-sensor systems and air defence systems. More recently, he has been the Principal Designer for the Thales WATCHKEEPER UAV system. He has been awarded the Royal Academy of Engineering silver medal in recognition of his contribution to radar, ISTAR and UAV systems. He is presently Technical Director for the Defence Mission Systems Business of Thales UK, with particular responsibilities for Radar, Electronic Warfare and Unmanned Air Systems.