This article appeared in the January/February 2019 issue of ADBR.

Machine learning, next-gen jammers to keep the EA-18G
at the forefront of electronic attack

At just US$7 million, the US Department of Defence contract awarded to Northrop Grumman for the Reactive Electronic Attack Measures (REAM) program last April seems pretty small beer.

Seven million dollars doesn’t go far in the world of defence procurement – maybe a dozen AIM-9X Sidewinder missiles, or not even a tenth of an F-35 Joint Strike Fighter.

But the US$7.26 million (about A$10.2 million) investment by the US Navy in REAM should result in a step-change increase in capability for the EA-18G Growler, the electronic attack aircraft now at the heart of the RAAF’s electromagnetic spectrum capabilities. Because the REAM program is funding the development of ‘machine learning algorithms’ – ie the use of artificial intelligence (AI) onboard the Growler.

“The REAM program is a future naval capabilities enabling capability with the objective of transitioning MLAs to the EA-18G airborne electronic attack suite to achieve capabilities against agile, adaptive, and unknown hostile radars or radar modes,” the US DoD contract award for REAM reads.

A “future capabilities enabling capability” is a particularly beautiful piece of military bureaucratic prose, but giving the EA-18G Growler machine learning will be central to its success in operating in a contested electronic warfare spectrum.

The US military has just declared the electronic magnetic spectrum a “domain”, explains John ‘JJ’ Thompson, Northrop Grumman Mission Systems’ NAVAIR campaign director. NAVAIR is the US Navy’s Air Systems Command, responsible for procuring and supporting US Navy (and Marine Corps) aircraft and weapons, including the Growler (and the F/A-18E/F Super Hornet it is based upon). Boeing is the EA-18G prime contractor, but Northrop Grumman is responsible for building and supporting much of the aircraft, from its fuselage to its jamming pods.

“Most of us can think about opposing armies manoeuvring throughout the ground terrain,” Thompson told Australian defence media in the US in January. “I’m going to be on a ridge line, just off a ridge line, I’ll be in a valley – all of those things in that physical space cause or restrict or give me advantage.

“The same thing applies to the electromagnetic spectrum. It is an advantage for me to own that space, because that allows me to communicate, to sense the world around me, and then to defend myself. I have to fight and achieve that space, and then work to dominate that space.”

And dominating that space means effectively detecting, classifying, and suppressing or destroying an adversary’s sensors and systems, such as radars and communications networks. The task, says Thompson, is to successful build your own ‘kill chain’, while breaking your adversary’s.

Military forces use the electromagnetic spectrum to build kill chains, Thompson, a former US Navy EA-6B Prowler electronic warfare officer and squadron commanding officer explains.

“Meaning, my ability to sense the environment around me, and then understand how do I put a weapon on that. That weapon can be kinetic or non-kinetic. How do I find that? I have to find, fix, track, and then engage. You have to build that kill chain.

“The same thing on defence. I have to break the kill chains of my enemies. So, if you’re trying to target me, I have to break it so that you can’t find me, can’t target me – all of these things I want to interrupt as we go back and forth. It’s an active conversation or battle, if you would, that this airframe (the Growler) plays heavily in.”

The challenge being increasingly faced in the electromagnetic spectrum today is that so-called ‘near peer’ actor that the US and its allies like Australia might face in a future conflict – ie a Russia or a China – have increasingly sophisticated radars and electronic attack capabilities of their own.

“This whole idea of how you sense, act and move with speed is where we’ve been investing really heavily inside the Growler, and looking at where Growler is going next,” Thompson said.

And to do that, correctly finding, identifying and classifying threats is critical.

It is here where the REAM program and machine learning comes into play, better allowing Growler crews to detect and identify different actors, assessing adversary radars and communication systems which may be rapidly changing their waveforms in real-time in an effort to confuse systems like the Growler.

“It’s really a (computer) processing race that we’re in. That’s where cognition really helps aircrews out now. It’s one individual engaging the screen and how much can my machine help me get to actionable things. There’s so many voices that are out there,” Thompson says.

Cognition begins to see patterns, Thompson explains, whereas in traditional electronic warfare, if an emitter can’t be matched against a known threat, it can be inadvertently disregarded. “I know that you used to work here. You seem very similar. Your pitch is the same. These other parameters are different. If your pitch is exactly like it was over here.” 

Machine learning allows the aircraft to begin assessing an emitting system’s intent. 

“I now can start to look for intent,” he said. “I go ‘Hmm, you seem like a long-range radar. You are located in a region that is not friendly to me. I’m going to start watching you’. Then, through the cognition, ‘I will watch your behaviour’. I’ve got all my ISR data cranking away going, ‘Does this guy match anything that I’ve ever seen before? Yes, No. Does he have attributes that are similar to what I’ve observed in the world, in this hostile place? Yes? No?’ And that’s how we deal with that.”

As well as machine learning, Northrop Grumman is working on another capacity to help the Growler detect and process threats – an expendable drone. Developed in partnership with Northrop Grumman by VX Aerospace, the Dash X drone, which was first test flown in late 2017 can be folded to fit in a 40cm diameter tube, such as a cluster bomb dispenser.

“This capability allows Dash X to be placed in a Tactical Munitions Dispenser (TMD) and flown on the wing of a tactical aircraft,” the VX Aerospace website explains. “This enables Dash X to be carried beyond the range of a similarly sized UAV before air deployment from the TMD, thus allowing Dash X to perform missions previously unachievable.”

Fitted with appropriate sensors, the 3.66m long Dash X cloud-operate in a manned-unmanned teaming arrangement with a Growler.

“How can we have a ‘friend’ of ours, very forward, loiter for an extremely long time inside a very hostile place, and can be handed off ship-to-ship? ‘We don’t ever want to see you again’. It’s a low cost enough that you can be attrited,” Thompson notes.

Back in December 2017 Thompson told US media that Northrop Grumman flew a demonstration where the Dash X “flew forward, looked for an unlocated RF [radio frequency] object, they went out and they found that vehicle, they listened for the whispering and they pulled it back to this test airframe [a Northrop Grumman-owned DHC-8 testbed] and they were able to detect, identify and geolocate.”

Identifying and classifying threats, and equally being able to sift through the electronic noise to determine spoof systems and what isn’t a threat is central to the Growler’s ability to then attack a target either ‘kinetically’ – ie launching a weapon against it – or non-kinetically through denying the adversary’s use of a radar or another emitter via jamming.

Today the Growler jams radars and other emitters with its ALQ-99 Tactical Jamming System pods, developed first for the EA-6B Prowler. Originally manufactured by EDO Corporation, now Harris Corporation, Northrop Grumman has been responsible for upgrading and maintaining the ALQ-99, in both its low-band and high-band forms since the late 1990s.

But the ALQ-99 pods are ageing and are set to be replaced under the US Navy’s Next Gen Jammer (NGJ) program, which ultimately will see, three, not two jamming pods procured, covering the mid, low and high-bands.

Work on the mid-band jammer – NGJ-MB – is well advanced after Raytheon won a competitive evaluation for its development in 2013. The NGJ-MB, designated ALQ-249(V)1, is currently in the engineering, manufacturing and development (EMD) phase, and should achieve initial operating capability (IOC) in 2022. 

Since late 2017 Australia has been a cooperative partner in the NGJ-MB’s development after the RAAF signed a memorandum of understanding with the US Navy in 2017.

Meanwhile, the NGJ low-band, or -LB, will be selected after a competitive evaluation of designs from L3 and Northrop Grumman. Both companies won US$35 million contracts last October to continue developing their pod designs, prototype examples of which will be evaluated by the US Navy.

One advantage Northrop Grumman is keen to champion of its NGB-LB contender is its intimate knowledge of the Growler airframe.

Like the high-band ALQ-99 pods, typically two NGJ-MB pods will be carried on the Growlers inboard underwing external stores stations (Stations 4 and 8). Meanwhile, the NGB-LB will be carried on the Station 6 centre-fuselage station.

This presents particular design challenges – the pod has to meet size and weight parameters, and has to be able to generate sufficient power to meet its own requirements without creating too big a drag penalty that reduces the Growler’s range. Like the ALQ-99 the NGJ pods will generate their own power rather than draw power from the host aeroplane.

“The airframe is designed to carry a certain amount of weight, [and] land [with a] certain amount of weight,” JJ Thompson explains. “Drag is also key to this. I could make something terribly inefficient and I can’t pour gas in the plane fast enough to move it through the space.”

Like the mid-band, the low-band pod can be expected to feature AESA technology. But given a competitive process is underway, neither Northrop Grumman nor L3 has given much detail away about their proposed designs.

But with Northrop Grumman responsible for designing and building the Growler’s centre fuselage, Thompson says that gives the company a competitive advantage in designing the most efficient NGJ-LB pod possible. “We have the advantage of actually having a fellow that did the aero design work,” he said. “They don’t get to sit with the guy that actually did the aero.”

Clearly the technical performance of the NGB-LB will be a key determinant in the winning design’s selection, but pod drag is vital, as a US Government Accountability Office decision denying a protest from Raytheon as to why its own NGB-LB design was not down-selected shows.

The redacted GAO report records; “The Navy explains that Raytheon’s proposed low band pod is approximately [DELETED] per cent larger than the ALQ-99 in the cross-sectional area. As such, the pod would significantly increase drag.”

One proposed design upgrade for the Growler that would increase range (by reducing drag) are the conformal fuel tanks Boeing has proposed for the Super Hornet and Growler airframes. The US Navy is now funding development of the tanks, but at this stage for the Super Hornet, and not the Growler. 

But given the Growler airframe is common with the Super Hornet, other than specific antenna apertures and other lumps and bumps, bringing the conformal tanks (which have been designed to be retrofittable to existing aircraft) across to the EA-18G should be relatively straightforward.

The conformal tanks, which would also be built by Northrop Grumman, would have a little less capacity than the large underwing fuel tanks they would be fitted in lieu of, but increase range thanks to generating less drag.

Conformal fuel tanks on the Growler would also improve the field of regard for the aircraft’s jammer pods, as underwing fuel tanks can partially block their electronic line-of-sight.  

Further capability enhancements for the Growler will come with the eventual development of a high-band NGJ (NGJ-HB). These would be much smaller pods carried on the Growler’s outer underwing weapon stations (stations 2 and 10).

Today those stations are typically used to carry the Growler’s two anti-radiation missile options, the AGM-88 HARM and the AGM-88E AARGM, although the mix of pods and weapons on the Growler depends on the mission.

Interestingly the Growler’s AARGM missile too is now a Northrop Grumman product, after the company acquired Orbital ATK in 2018. What was Orbital ATK, which builds rockets, rocket motors, satellites, missiles and small arms, is now known as Northrop Grumman Innovation Systems.

The AARGM (Advanced Anti-Radiation Guided Missile) is a rebuilt AGM-88 HARM, which Northrop Grumman disassembles and then rebuilds with a new guidance system featuring a millimetre wave radar seeker and GPS/INS, while the missile control section is substantially modified.

The RAAF has acquired limited stocks of both AARGMs and HARMs for its now 11-strong Growler fleet. Likewise, given its desire to remain in lockstep with the US Navy’s development pathway for the jet, it can be expected in due course to acquire the Next Gen Jammers, upgrade them with the machine learning capabilities and fit conformal tanks, assuming they are migrated across to the EA-18G platform.

Like the electromagnetic domain it operates in, the Growler will continue to evolve.