Agnostic Electronic Warfare
Agnostic Electronic Warfare (EW) is a school of EW theory emerging from current and expected technological advances in the Artificial Intelligence (AI) and Machine Learning (ML) domains.
To an extent, Agnostic EW turns conventional EW theory on its head.
The latter prizes the centrality of the electronic support element of the EW theory triumvirate as the foundation for the pursuit of electronic attack and electronic protection. Traditionally it has been difficult, if not impossible, to perform electronic attack and electronic protection without electronic support.
Electronic support is facilitated by the collection of Signals Intelligence (SIGINT).
SIGINT includes Communications Intelligence (COMINT) on friendly and hostile communications and raw Electronic Intelligence (ELINT) concerning friendly and hostile radars.
Analysing SIGINT enables EW practitioners to detect, characterise and locate blue and red forces, and unknown emitters in their area of responsibility.
This analysis facilitates the drafting of an Electronic Order of Battle (EORBAT).
The EORBAT will identify these emitters, reveal their locations and provide additional information like emitter frequencies and waveforms.
The EORBAT is not fixed, changing regularly during manoeuvre as the battlespace ebbs and flows.
Since the dawn of electronic warfare during the Second World War the EORBAT has been the essential ‘sheet music’ for the prosecution of EW at the operational and tactical levels.
It informs personnel and platforms which emitters to attack kinetically or electronically, which to leave alone lest they be friendly or exploited by blue force analysts for SIGINT, or which are to be physically avoided, particularly relevant for radar.
The electronic order of battle is used to programme the threat libraries of electronic warfare systems, particularly their ESMs.
This information allows the ESMs to recognise hostile emitters and to alert personnel and/or trigger a response.
For example, programming an ESM with the most relevant emitter information ensures the electronic attack systems can engage a hostile radar with the most efficient jamming waveforms.
Likewise, EORBAT information can be briefed to personnel prior to their mission to highlight electromagnetic threats they may encounter, where these threats are and what action can be taken to reduce their efficacy.
Agnostic EW allows the prosecution of electronic attack and protection by using either a very rudimentary EORBAT which may, at best, have scant information regarding blue or red force emitters, or without an EORBAT altogether.
Although still a theoretical proposition, agnostic EW could exploit technology to facilitate its emergence in the coming years.
It is an aphorism of electronic warfare that electronic attack and protection systems are only as capable as their threat information.
If EORBAT information is fragmented, corrupted or inaccurate then the ability of an electronic attack/protection system to respond to a threat may be degraded to a point of ineffectiveness.
At the start of the 21st Century’s second decade electronic warfare writ large is witnessing two major advancements:
- The first is the advent of artificial intelligence and machine learning.
- The second is the proliferation of wideband Internet Protocol (IP) communications across the battlespace.
This facilitates the movement of situational awareness and command and control information across Line of Sight (LOS) and Beyond LOS distances using conventional and satellite communications.
Advances in AI/ML and wideband networking are integral to the advent of Cognitive Electronic Warfare, of which agnostic EW is a subset.
Cognitive EW systems use information amassed during their use to sense and comprehend the electromagnetic environment.
An ESM equipping a combat aircraft employing cognitive techniques might have amassed terabytes of information regarding the ground-based air surveillance, naval surveillance and fire control/ground-controlled interception radars it has encountered during its service life. AI/ML software equipping the ESM will allow it to recognise specific radars and to trigger the aircraft’s electronic attack system to transmit a particular jamming waveform into a particular radar.
The choice of waveform employed may result from an evaluation of which waveforms have been successful in the past against this particular radar.
In a nutshell cognitive EW aims for electronic warfare equipment to recognise and understand what they are seeing and to inform the human operator thus, or to initiate a response with a modicum, or with zero, human intervention.
The advent of advanced wideband communications in the battlespace assists EW systems in exploiting cognitive approaches by continually updating these systems with new data relevant to their mission while sharing their data with other off board platforms, sensors and effectors.
This networking will allow the decision-making capabilities of cognitive-enabled electronic warfare systems to improve continually as they receive real-time updates on the electromagnetic environment.
The logic is that the more information an EW system receives, both before and during its mission, the better its response to electromagnetic threats.
While cognitive EW covers the whole gamut of strategic, operational and tactical electronic warfare agnostic EW is focused on the tactical dimension.
The key tenet of agnostic EW is to employ AI/ML approaches to enable tactical electronic warfare systems to immediately recognise a potential electromagnetic threat and to either take a course of action against that threat autonomously, or to recommend a course of action to a human operator despite having a minimum of pre-programmed threat information.
The so-called ‘holy grail’ of this approach would be for an ESM to detect a transmission and to use cognitive algorithms to immediately recognise the waveform and to then decide one or more courses of action as a response based upon responses which have been successful in the past.
It is not that an agnostic EW system would require no programming.
The equipment would need the correct AI/ML machine learning algorithms and perhaps a basic library detailing the characteristics of the threats it must recognise.
The intention is for the AI/ML software to help the system learn, respond and evaluate as it moves through its service life.
The goal of an agnostic EW system should be to avoid having to load and re-programme threat libraries prior to each mission.
Instead, the AI/ML software will enlarge and enrich this data.
Networking an agnostic EW system with other similar systems in the battlespace will allow them to send and receive emitter data as and when encountered, along with details of responses as and when performed.
This will help to enrich continually in real time not only the quality of the data that an agnostic EW system works with, but also the quality of its decision-making.
Like the advent of cognitive EW writ large, the dawn of agnostic electronic warfare is not a question of if, but when.
Much of the technology required to realise true agnostic EW capabilities still needs invention, yet AI/ML approaches are penetrating the wider military domain and so they will penetrate the world of EW. There is an inevitability driving the advent of agnostic EW.
Software-defined radars and radios are themselves adopting cognitive techniques.
This will result in an exponential increase in the variety of waveforms these devices will transmit.
Such will be this profusion that it may eclipse the abilities of the human brain to comprehend and address.
Agnostic EW systems will be effective at the tactical edge by enabling electronic attack, protection and support without necessarily having a bedrock of data regarding the emitters that electronic warfare apparatus may face in a particular theatre.
Instead, a priori approaches using AI/ML could help these systems recognise and address emitter threats as and when encountered.
Quite simply, we may face a future where electronic warfare equipment can no longer be programmed for each and every scenario it might face.
This will be amplified by operational and tactical realities helping to drive the adoption of agnostic approaches in EW.
Armed forces may receive little or no warning of the conflicts in which they become involved.
This may preclude opportunities to collect relevant SIGINT on hostile emitters.
Such ‘come as you are’ conflicts will need platforms and subsystems with embedded agnostic EW capabilities to ensure that hostile emitters can still be recognised and engaged even if red force SIGINT is sparse.
Agnostic EW will become a reality on the battlefield over the coming decades.
The ever-growing complexity of the electromagnetic environment will help drive its adoption as will advances in artificial intelligence and machine learning.
Despite primarily being applicable to the tactical level of war, agnostic EW will have a major role to play in enhancing platform and personnel protection against electromagnetic threats which are becoming similarly complicated.