Trouble in the spectrum

An innovative, new system employing artificial intelligence (AI) could help enhance the resilience of US Army tactical networks on the battlefield against aggressive electronic warfare.

“Shoot, move, communicate” was the mantra of the US Army’s tactical doctrine in the 1980s. Seen as the three basic tasks a solider must achieve to not only survive, but prevail in battle, it has stood the test of time. The advice is as relevant today as it was 40 years ago. That said, the latter stipulation is as likely to be as harshly challenged by the enemy today as the first two.

During the Cold War, US land forces and their NATO counterparts were cognisant of the power of Warsaw Pact electronic warfare (EW). If NATO and Warsaw Pact forces had come to blows across the Inner German Border (IGB), the ether would have been thick with radiation. Electromagnetic waves in the form of ionising radiation would have endangered life and limb as the result of nuclear explosions. Meanwhile, Warsaw Pact EW cadres would have listened to the airwaves to detect and locate NATO land forces’ radio communications. Once discovered, radios used by soldiers, vehicles, bases, weapons and aircraft would have been blasted with Warsaw Pact jamming. Attacking NATO radio communications was imperative as Warsaw Pact spectrum warriors would work to degrade, damage and destroy the links land forces relied on for command and control (C2). Some tactical/operational links would no doubt have been left unjammed to be exploited for intelligence. Pinpointing NATO units based on their radio emissions would provide aimpoints for fires and close air support. All-in-all, the electromagnetic spectrum in and around the IGB, and European theatre of operations, would be an arena of complete chaos.

The US Army’s mantra today holds an added sense of déjà vu. “Gentlemen, after a nice little vacation, looks like we’re back at it again,” said Captain Franklin ‘Frank’ Ramsey in the 1995 thriller Crimson Tide. Capt. Ramsey, played by Gene Hackman, was commenting on renewed US–Russia nuclear tensions. In the film, Russia had descended into civil war and her nuclear C2 had been compromised. Art can imitate life and NATO finds itself once again locked in tension with Russia in a new era of geopolitical rivalry not unlike the Cold War.

During 2008, in an attempt to arrest years of post-Cold War decline, and in the wake of a short war with Georgia, the Russian government launched what were termed the ‘New Look’ defence reforms. For Russia’s land forces, this overhauled the army’s order-of-battle. Russian land forces also include naval infantry and airborne forces, each of which are separate services. Changes in orders-of-battle were matched with materiel modernisation. It did not go unobserved that Russian land forces’ EW capabilities have received significant investment, with the service entry of a host of new capabilities. In 2014, during the first Russian invasion of Ukraine, several of these new EW systems deployed to the theatre of operations where they demonstrated their efficacy against Ukrainian radio communications.

Despite a lacklustre showing at the start of their second invasion of Ukraine in February 2022, Russian EW practitioners have learned from their mistakes. An article published in January 2024 in the Financial Times entitled ‘Russia has the upper hand in electronic warfare with Ukraine’ pulled no punches. Russian electronic jamming remains potent and is particularly effective against Ukrainian uninhabited aerial vehicles (UAVs). Anecdotal evidence shared with the author highlights some of the problems: Unencrypted radio links connecting the aircraft to their pilot for C2 and telemetry have been vulnerable to electronic attack. Likewise, global navigation satellite system (GNSS) position, navigation and timing (PNT) radio signals – that UAVs rely on for navigation – have also been at risk. GNSS PNT signals used by precision-guided weapons such as US-supplied Boeing joint direct attack munitions (JDAMs) have been adversely affected by Russian jamming. PNT vulnerabilities were highlighted in a trove of classified US Department of Defense (DoD) documents, leaked in 2023.

Moreover, Russian EW has proven effective against Ukrainian tactical communications lacking in robust communications/transmission security (COMSEC/TRANSEC) protocols. Techniques including encryption can help protect traffic from eavesdropping. Frequency hopping can also frustrate an opponent’s ability to detect the radio transmissions in the first place and then jam them. This is because the signal’s frequency is continually changing in a pseudorandom fashion, sometimes several thousand times per second. However, one seemingly insurmountable problem is that whenever forces emit on the battlefield, they announce their presence. Land forces may have sophisticated radios with a myriad of COMSEC/TRANSEC techniques, and low probability of detection/interception techniques may try to keep signals as discreet as possible; nevertheless, every radio transmission must move through the ether.

Pathfinder

News came to light in November 2023 that the US Army had deployed a new capability called the Advanced Dynamic Spectrum Reconnaissance (ADSR) system during a multinational exercise in Germany. Reports said that ADSR uses artificial intelligence (AI) techniques to let deployed radio networks detect, and then avoid, electronic attack.

ADSR was developed under the US Army Research Laboratory’s (ARL’s) Pathfinder initiative. Pathfinder was launched in June 2021 to harness academic know-how and expertise to rapidly solve problems the US Army is facing. The programme is managed by the Armaments Centre of the Army’s Combat Capabilities Development Command (DEVCOM) in conjunction with the ARL. Both organisations are in turn working with the US Army’s 18th Airborne Corps’ 82nd and 101st Airborne Divisions. Academic assistance has been provided by universities in North Carolina, Tennessee and West Virginia. The universities have been supported by the US Defence Advanced Research Projects Agency in their ADSR endeavours.

How does ADSR work? Reports covering the recent deployment of the system to support an exercise in Germany involving the 101st Airborne Division provides some clues. The exercise took place in late 2023 at the US Joint Multinational Readiness Centre in Bavaria, southwest Germany. Essentially, ADSR exploits the Army’s own deployed tactical networks to achieve two tasks: First, ADSR works to reduce tactical radio frequency (RF) emissions writ large across the battlefield. Second, the system also employs these networks to sense and avoid hostile jamming. The logic here is two-fold; prevent networks, radios and hence assets (personnel, vehicles, bases, weapons, sensors and capabilities) being detected via their RF emissions. What has not been detected cannot be jammed. Likewise, by ascertaining where hostile jamming is occurring, areas where jamming may be prevalent can be avoided.

Networks

US Army units deploy a bewildering array of tactical networks on the battlefield. The force uses TrellisWare’s TSM very/ultra-high frequency (V/UHF: 30 MHz–3 GHz) waveform. TSM replaces the US Army’s erstwhile Soldier Radio Waveform (SRW), a UHF waveform for intra-platoon and company communications. Moving up in echelon, company headquarters use the Wideband Networking Waveform (WNW) and Army Networking Waveform-2 (ANW2). The WNW uses V/UHF frequencies, while the ANW-2 is restricted to UHF (300 MHz–3 GHz). Both waveforms carry tactical voice and data traffic between vehicles, deployed headquarters and dismounted troops. Company-level command posts can also access TSM networks. WNW networks, meanwhile, connect company headquarters to their battalion-level counterparts. A plethora of satellite communications (SATCOM) constellations provide beyond line-of-sight (BLOS) links to a US Army deployed manoeuvre force.

Since the commencement of the US-led counter-insurgency operation in Afghanistan and Iraq just after the turn of the century, the US Army has primarily organised its manoeuvre force around the Brigade Combat Team (BCT). This is now changing with the force adopting a divisional structure. According to the US DoD, the reorganisation around larger formations is to enable overmatch against near-peer rivals. While not named explicitly, these rivals are understood to be the People’s Republic of China and Russia. The advent of the divisional structure should not have too big an impact on current manoeuvre force communications. Nonetheless, the reorganisation could see a higher reliance on BLOS links such as SATCOM. Two main BLOS networks are used by the manoeuvre force: The Warfighter Information Network-Tactical (WIN-T) is joined by the Mobile User Objective System (MUOS). Both MUOS and WIN-T use V/UHF links and are primarily used at battalion and company levels. Units and headquarters at company level and below use SATCOM networks provided via the Integrated Waveform (IW) and the JBCP (Joint Battle Command Post). The IW is a UHF waveform with the JBCP, which is the Army’s blue force tracking system, also using UHF.

As one can see, the manoeuvre force relies on a plethora of networks to maintain communications; a profusion which is not accidental, and which provides advantages from a redundancy perspective. Successful electronic attack against one or two of these networks will not deprive the manoeuvre force of communications. However, as the war in Ukraine has illustrated, Russian land forces take electronic attack very seriously. The Russian Army deploys three systems at the tactical level to detect, locate and jam V/UHF radios and networks. These EW platforms include the R-330B Borisoglebsk-2, R-330Zh Zhitel and RP-377U/UV. Ukrainian sources have shared with the author that encrypted waveforms have remained robust in the face of severe Russian jamming. US-supplied Single Channel Ground and Airborne Radio System (SINCGARS) transceivers have held their own when bombarded by Russian electromagnetic waves; this resilience is made more remarkable by the fact that this radio system relies on a design over 40 years old. Nevertheless, it would be negligent to rely on waveform COMSEC/TRANSEC to act as the first and last line of defence against jamming.

Listening to the ether

The stakeholders involved in ADSR declined to publicly share further information on the system. As a result, one must resort to speculation to understand how ADSR might work. As articulated above, ADSR uses deployed army networks to sense and avoid jamming. This means that ADSR must have some means by which to ascertain what is happening in the ether.

Situational awareness to this end could be delivered via deployed US Army EW systems. The force is in the process of receiving new manoeuvre force EW platforms. These platforms are built around the Army’s Terrestrial Layer System (TLS) which constitutes two distinct capabilities. Tactical land manoeuvre force EW will be supported by the TLS Brigade Combat Team (TLS-BCT) ensemble. Electronic warfare at the operational level will be performed by the TLS Echelon Above Brigade (TLS-EAB) system. Lockheed Martin is currently developing TLS-EAB and TLS-BCT prototypes. According to the US Army, the force is expected to complete the introduction of both the TLS-BCT and TLS-EAB between 2030 and 2035. The two TLS configurations will be joined by a backpack electronic attack system intended for dismounted troops. Mastodon Design, part of CACI International, won a USD 1.5 million contract to provide a prototype in late 2023. The backpack forms part of the TLS-BCT architecture.

Manoeuvre force EW elements are knitted into Raytheon’s Electronic Warfare Planning and Management Tool (EWPMT). The Army says the EWPMT provides electronic warfare C2 and training, in support of electromagnetic manoeuvre. In short, the EWPMT acts as the clearing house for incoming Signals Intelligence (SIGINT) and subsequent outgoing electronic and cyberattack taskings. One concept of operations for ADSR could be for it to act upon information collated by the EWPMT. Let us suppose that a BCT’s infantry battalion is experiencing jamming in its area. The jamming has been detected by the BCT’s organic TLS-BCT systems. The TLS-BCT has determined the areas being most adversely affected by the jamming. Red force electronic attacks are degrading parts of the tactical networks used by the battalion in the affected areas. ADSR could receive notifications from the EWPMT regarding these affected areas. Working with network management software, ADSR could present options for configuring these networks to avoid jamming. One option presented by ADSR could be to alter the network’s topology. This might mean that traffic from the affected infantry battalion’s units follows paths via nodes further back from the tactical edge to avoid the worst of the jamming.

Similarly, ADSR may be connected with the network’s management software and continually monitor the network’s performance. By using machine learning approaches, ADSR’s software could be trained to recognise when jamming is taking place. For instance, if traffic suddenly becomes intermittent or stops altogether on one part of the network, this may indicate that jamming is occurring. Available bandwidths suddenly experiencing significant constrictions may provide an additional, similar clue. In a sense, tactical communications mobile ad-hoc networking (MANET) approaches have some of this functionality built-in. If part of the network is compromised for whatever reason, the network reconfigures to continue functioning. Fusing MANET approaches with those of ADSR could increase and deepen network integrity. Using tactical networks to sense and react to jamming enhances the force’s overall spectrum manoeuvre capabilities. It is possible that ADSR can supplement the TLS-BCT or even assume some of its SIGINT burden.

ADSR is also tasked with reducing tactical communications emissions across the battlefield. How this might work in practice is less clear. For example, ADSR may continually monitor RF emissions via the network’s management software. Once again, machine learning may have much to offer. ADSR software could be trained to understand an infantry battalion’s usual radio emission behaviour at various stages of battle. The software could correlate this behaviour with effective jamming incidents. If the battalion’s emissions were at a particular level, did the enemy start jamming? Was this because the strength of the battalion’s radio signals were at a level that could be detected with relative ease by red force SIGINT? Once detected, how severe was red force jamming? How long did the jamming last? Where was the jamming concentrated and how effective was it against friendly emitters? There are a myriad of factors that ADSR software could account for to determine when blue force emissions prompt a red force response. All these factors could help ADSR’s algorithms advise how the network should be configured to remain survivable. Recommendations could be shared with the network’s management software which can make the necessary alterations to reduce emissions.

Challenges ahead

It is important to remember that AI is not a silver bullet. Like all aspects of computing, it depends on the reliability and quantity of data that it can be trained with. ‘Garbage in, garbage out’ (GIGO) is an oft-used refrain, but it may be a paucity of data that capabilities such as ADSR will have to address. Mercifully, the US and its allies have not found themselves embroiled in conflicts involving a peer- or near-peer adversaries in recent years. Those wars that have involved these actors over the past 30 years have tended to feature low-tech opponents. Enemies such as these were unlikely to deploy sophisticated jamming against allied land forces’ communications networks. A lack of ‘real world’ data will force ADSR’s AI techniques to be trained with simulated data. Additional useful information may be culled from signals intelligence data shared by the Ukrainian military with its allies. A significant ongoing SIGINT ‘soak’ of the Ukrainian theatre of operations by the US and others may be helpful in this regard. At the very least, ADSR will have a reservoir of data that can be used to train its algorithms.

The reticence of the US Army and associated stakeholders to discuss ADSR make it difficult to articulate the system’s capabilities with any certainty. Nonetheless, by examining information in the public domain and combining this with educated conjecture, one can contemplate the system’s capabilities. ADSR’s realisation comes at an opportune moment. The US Army will have to fight hard to win and retain control of the spectrum as a manoeuvre space in future conflicts. ADSR will have an important contribution to make in fulfilling this mission.

Thomas Withington