Network Centric Warfare (NCW) is the buzzword of choice in current Defence Department rhetoric. There is little doubt that the introduction of NCW is the defining paradigm of this decade in military affairs, and inevitably, we should see this reflected in Australia. How well NCW is understood in Australia's DoD is, however, another matter entirely.

From a broad perspective the introduction of networking techniques into warfighting systems is the military equivalent of the digitisation and networking drive we observed in Western economies between 1985 and 1995. Military networking, especially between platforms, is far more challenging than industry networking due to the heavy reliance on wireless communications, high demand for security, and the need for resistance to hostile jamming. The demanding environmental requirements for military networking hardware are an issue in their own right. It should come thus as no surprise that the introduction of networking into military environments has proven more painful and more protracted than the industry experience of over a decade ago.

Why Networking?

Much has been written and said over the last decade as to why networking is essential, and how it improves warfighting capability. Unhappily, not all of the publications present robust arguments, and much of what has been written has been accepted as fact, rather than critically analysed.

At the most fundamental level networking aims to accelerate engagement cycles and operational tempo at all levels of a warfighting system. This is acheived by providing a mechanism to rapidly gather and distribute targeting information, and rapidly issue directives. A high speed network permits error free transmission in a fraction of the time required for voice transmission, and permits transfer of a wide range of data formats.

In a more technical sense, networking improves operational tempo (optempo) by accelerating the Observation-Orientation phases of Boyd's Observation-Orientation-Decision-Action (OODA) loop. Identified during the 1970s by US Air Force strategist John Boyd, the OODA is an abstraction which describes the sequence of events whihc must take place in any military engagement. The opponnent must be observed to gather information, the attacker must orient himself to the situation or context, then decide and act accordingly. The OODA loop is thus fundamental to all military operations, from strategic down to individual combat. It loop is an inevitable part of reality and has been so since the first tribal wars of 25,000 years ago, as it is fundamental to any predator-prey interaction in the biological world. Sadly, its proper understanding had to wait until the 1970s.

At a philosophical and practical level what confers a key advantage in engagements is the ability to stay ahead of an opponent and dictate the tempo of the engagement - to maintain the initiative and keep an opponent off balance. In effect, the attacker forces his opponent into a reactive posture and denies the opponent any opportunity to drive the engagement to an advantage. The player with the faster OODA loop, all else being equal, will defeat the opponent with the slower OODA loop by blocking or pre-empting any move the opponent with the slower OODA loop attempts to make.

The four components of the OODA loop can be split into three which are associated with processing information, and one which is associated with movement and application of firepower. Observation-Orientation-Decision are information centric while Action is kinematic or centred in movement, position and firepower.

If we aim to accelerate our OODA loops to achieve higher operational tempo than an enemy, we have to accelerate all four components of the loop. Much of twentieth century warfighting technique and technology dealt with accelerating the kinetic portion of the OODA loop. Mobility, precision and firepower increases were the result of this evolution.

There are practical limits as to how far we can push the kinetic aspect of the OODA loop - more destructive weapons produce collateral damage, faster platforms and weapons incur ever increasing costs. Accordingly we have seen evolution slow down in this domain since the 1960s. Many weapons and platforms widely used today were designed in the 1950s may remain in use for decades to come, the B-52 being a good case study.

Observation-Orientation-Decision are all about gathering information, distributing information, analysing information, understanding information and deciding how to act upon this information. The faster we can gather, distribute, analyse, understand information, the faster we can decide, and arguably the better we can decide how and when to act in combat. Networking is a mechanism via which the Observation-Orientation phases of the loop can be accelerated, and the Decision phase facilitated.

Well implemented networking can contribute to improved effectiveness in other ways. One such technique is 'self synchronisation' which permits 'directive control'. Rather than micromanage a warfighting asset with close control via a command link tether, warfighters are given significant autonomy, defined objectives, and allowed to take the initiative in how they meet these objectives. A fighter pilot who receives continuous updates from an AEW&C aircraft over a network can make his own tactical decisions, exploiting the situational picture broadcast from the AEW&C aircraft. This is of course not a new model, but networking facilitates it in ways difficult to match - compare this example with the AM radio broadcasts issued in 1943-1945 by Luftwaffe air defence nodes to permit night fighters to autonomously hunt for Allied bombers. Both amount to a 'self synchronisation' scheme, but the underlying technology is five decades apart.

Networking is not a panacea, nor can it be. The ultimate limits on the combat effect produced by a warfighting system, and thus is capability, are bounded by the Action or 'kinetic' phase of the loop. Bombs or missiles delivered is the bottom line, and networking is a tool to facilitate this effect, it is not a sustitute for bombs and missiles on target as some proponents of NCW publicly advocate.

Until recently the mathematics underpinning capability gains in large networked warfighting systems had not been studied closely. Many scholars of NCW simply borrowed the well established Metcalfe's Law from the commercial domain and simply declared it to be valid for military systems. Metcalfe's Law states that the 'utility' of a network increases with the square of the number of nodes in the network - ten nodes (platforms) permit a hundred possible connections, a hundred nodes ten thousand. Unfortunately the mathematics of web site driven sales statistics are not particularly revelant to the behaviour of networked military forces. What we now understand is that Metcalfe's Law presents a possible best case scenario for distribution of information collected by sensors on platforms in a military system. At best it is an indicator of gains in situational awareness, assuming the data being distributed is valid, timely and relevant. The real limits to capability gains in networked systems arise from the Decision-Action phases of the OODA loop.

The Decision phase sees a commander exploiting knowledge acquired in the Observation-Orientation phases, and conferring as required with his superiors and subordinates to determine what is the best choice of action. In the Action phase the commander must deploy his assets and effect the engagement. Both of these phases of the loop, in mathematical terms, are queuing systems. The commander must wait for others to respond, and must marshal and position assets to engage. All of these events involve one entity waiting for another, in effect queueing up.

The mathematical model which contrains such systems is Amdahl's Law, like Metcalfe's Law a defining equation in the computer industry. The reality Amdahl defines is simple - increasing the number of assets in the system increases the achieved work or effect at best only by the number of assets added. The actual improvement is limited by the queuing effects seen in marshalling and positioning assets to perform engagements.

The mathematical bottom line in NCW is a very simple one: networking can permit a significant improvement in operational tempo, where a shortage of targeting information is the bottleneck to achieving a high operational tempo, but networking itself has very little impact on the absolute ability of a force to deliver weapons against targets, that being constrained by the capabilities and number of combat platforms in use.

Networking can accelerate operational tempo by speeding up the Observation and Orientation phases of Boyd's OODA loop. Unfortunately the bounds on the capability of the 'system of systems' are imposed by the Decision and especially Action phases of the loop (Author).

A good example is the classic Desert Storm scenario of an air force attacking strategic targets and in situ battlefield targets like deployed armoured divisions. With ample targeting information, especially for fixed targets, networking of the attacking force would not have dramatically increased combat effect.

Accordingly we have seen networking produce its greatest gains in combat effect during battlefied strike and close air support operations, especially against highly mobile and fleeting ground targets. In such an environment, where the opponent is continuously on the move, networking can produce spectacular gains since the bottlneck limiting force capability lies in the flow of targeting information to strike aircraft.

No less interesting are the effects observed in demand for specific types of assets to support networked interdiction and strike operations. Short targeting cycles in strike operations require that the bomber be orbiting in close proximity to the intended target - this is persistent strike formally now labelled a killbox interdiction. In practice this has driven up the demand for tanker sorties and the demand for B-52H, B-1B and F-15E - all at the expense of the smaller F-16C and F/A-18 variants. Bigger is better in the networked strike game, so much so that a recent discussion piece by US analyst Price Bingham in the ISR Journal predicted the demise of the classical battlefield interdiction tasked fighter-bomber, in favour of larger bombers and UCAVs. This is a direct challenge to the basic rationale for the Joint Strike Fighter family of battlefield interdiction and close air support fighters, and the longer term use of legacy designs like the F-16 and F/A-18 variants.

Networking has produced other useful benefits. One is combat identification and deconfliction, where the JTIDS/Link-16 system has been used very effectively as a more capable substitute for conventional IFF, capable of supporting air, land and sea assets.

A key issue for all networking is the Intelligence-Surveillance-Reconnaissance capability supporting it. Networks like all computing systems obey the Garbage-In Garbage-Out rule - without accurate high quality ISR systems feeding the network, it is little more than high speed digital plumbing between platforms, with nothing useful to carry. In the US force structure, the pressure to introduce network developed mostly due to bottlenecks in pushing ISR derived targeting information out of existing ISR systems like AWACS, JSTARS and Rivet Joint.

Networking capabilities are not confined to Western nations. The diffusion of commercial computing and networking systems globally has contributed to a growing in focus by non-Western militaries. Russia has capitalised on this by aggressively marketing ISR platforms like the A-50 AWACS, digital datalinking products - the Soviets were deeply enamoured of digital air defence networks - and counter ISR systems. The latter include long range AAMs like the R-172, R-37 and Kh-31 variants, as well as airborne and land mobile high power jamming equipment, and very long range SAMs like the S-400 and Imperator series.

In summary the introduction of networking offers many benefits for an air force and should be actively pursued. It is however not a substitute for combat or ISR capabilities in a force structure and cannot be used as an excuse to justify downsizing of combat fleets.

The Technological Perspective

The technology supporting NCW is inherently complex, but not significantly moreso than the technology used to digitise and network the civilian world. It must however be more resilient physically, thermally, electrically and be better resistent to hostile penetration, and in wireless systems, hostile jamming.

The prerequisite for an NCW capability is the digitisation of combat platforms. A combat aircraft with a digital weapon system can be seamlessly integrated in an NCW environment by providing digital wireless connections to other platforms. Without the digital weapon system, and its internal computers, NCW is not implementable. The growing gap between the US military and the EU military largely reflects the Europeans' reluctance to heavily invest in digitising their combat platforms.

Provision of digital wireless connectivity between combat platforms is a major technical challenge which cannot be understated. While civilian networking of computers can largely rely on cabled links, be they copper or optical fibres, with wireless connectivity as an adjunct, in a military environment centred in moving platforms and field deployed basing, wireless connectivity is the central means of carrying information.

The problems faced in providing military networking are generally well understood, but often push the boundaries of available technology. Key issues can be summarised thus:

At the time of writing nearly all military datalinks used in NCW operate at speeds which would be considered intolerable in the civilian/commercial world, reflecting the realities of wireless communications. Moreover, the military world lives with a veritable Tower of Babel in both signal modulations, operating frequencies and digital communications protocols, and variations of nominally standard protocols.

To contextualise this, Western armed forces currently deploy systems using a wide range of current and legacy signal formats and protocols, examples being:

This veritable menagerie of datalink modulations/protocols is by no means exhaustive, but reflects the realities observed in the computer industry in the decades predating the Internet. New protocols like the Joint Tactical Radio System (JTRS) are in part intended to incorporate mechanisms for translating such legacy protocols into formats which can be sent over a common channel. Separate from these multi-platform protocols and modulations are the type specific datalinks, such as the intra- and interflight datalinks used on the F/A-22A and later the JSF.

As yet there has been little effort to capitalise on the new technology of ad hoc network protocols, designed for self organising networks of mobile platforms, although the JTRS WNW effort looks promising. The DARPA GLOMO program in the late 1980s saw considerable seed money invested, but did not yield any publicised dramatic breakthroughs. Ad hoc networking remains a yet to be fully explored frontier in the networking domain, one which is apt to provide a decisive technology breakthrough for NCW.

The technological issues in NCW often dominate the debate at the expense of the deeper philosophical and functional issues which is unfortunate, since both domains matter and getting either wrong results in an equally disfunctional end result.

NCW and Australia

In Australia networking has been very much in the limelight of the defence debate. Sadly, it has also been used to justify a great many dubious decisions, all predicated on premises which do not hold.

Like most intellectually demanding and complex systems problems, NCW must be properly understood before it can be used as a basis for strategic planning decisions. Clearly this has not been the case in many key areas of the DoD, resulting in public statements which would be comical were not the circumstances so dire.

Perhaps the best exposition of this problem lies in the package of submissions presented earlier this year to the JSCFADT committee of Federal Parliament, especially the Air Combat Capability paper.

In this document we learn that networking can substitute for combat fleet numbers, despite the contrary experience in Afghanistan and Iraq, where aircraft size became the pivotal issue. We also learn that the combat effect of the system should exceed the sum of its parts despite the contrary mathematics underpinning this problem. Limitations in the F/A-18A and JSF we learn do not matter, since Australia will evidently have an asymmetric advantage in networking and AEW&C, despite regional buys of Russian/Israeli A-50 AWACS, and Russian marketing of TKS-2 digital datalinks on Su-30 fighters. We also learn in this document that AEW&C aircraft and networks will not be challenged, as evidently Russian R-172, Kh-31 and R-37 missiles, or high power jammers, will never be used in the region - regional buys notwithstanding. No differently, Australia need not invest in high power jamming aircraft since other nations will never use their AEW&C and networking systems. We also learn that five A330-200 tankers will improve the RAAF's combat persistence, since evidently the fuel carried in F-111s need not be counted. No less surprisingly we also discover that the mere AU$20M to 30M required to put networking into the F-111s is unusually expensive and not worth doing - evidently it is better to kill off the aircraft than invest into networking it.

The ugly reality is that networking has become a cure all panacea in the DoD bureacratic machine, one which can magically offset all force structure limitations, and one which magically only Australia can possess and use properly in the Pacrim.

The analytical perspective is very different, however. Most regional nations are now operating, deploying or shopping for AEW&C aircraft. Russia is actively marketing digital datalinks, like the TKS-2 and older APD-518, and marketing counter-ISR weapons like the Novator R-172 (KS-172) or Kh-31 series missiles. Russia is also marketing high power jamming equipment, especially pods using Digital RF Memory (DRFM) technology, and there is a good prospect of a Growler-ski based on the Su-32 materialising before the end of the decade.

In practical terms, by 2010-2015 regional opponents without AEW&C, long range counter-ISR missiles and jamming pods are likely to be the obliging exception to the rule. US thinking is not surprisingly centred in using F/A-22As to sanitise airspace permitting unhindered use of ISR platforms and networks, and the program to replace the lost capabilities of the EF-111A Raven with the B-52J or EB-52, equipped with high power stand-off jamming equipment to disrupt opposing networks and ISR sensors.