The Future of AEA. Also in this issue: Cognitive EW Technology Survey: Flightline and Portable EW Simulators 2016 AOC Award Winners

Transcription

1 DECEMBER 2016 Vol. 39, No The Journal of Electronic Defense The Future of AEA Also in this issue: Cognitive EW Technology Survey: Flightline and Portable EW Simulators 2016 AOC Award Winners

2 The Future of Airborn Building a Smarter IIt s already clear that US airborne forces and platforms will need to operate in much more highly-contested Anti Access/Area Denial (A2/AD) environments than they have ever faced before. This already daunting challenge is only expected to increase, and as a result, the Services airborne electronic attack (AEA) capabilities and overall AEA strategy will need to be dramatically improved and restructured to keep pace with the threat. The DOD has determined, however, that this requirement cannot be met with incremental improvements to existing systems, but 38 rather must represent an exponential advancement and implementation of the very latest technological capabilities, specifically artificial intelligence (AI) architectures and machine-learning techniques and algorithms. In the context of AEA, these technologies are described as cognitive EW self-learning systems combined in highly-integrated networks able to detect, react and adapt to threats in real-time. STAND-OFF TO STAND-IN The most pressing concern for AEA planners is the increased operating range and sophistication of advanced integrated air-defense systems (IADS) incorporating both active and passive surveillance and targeting capabilities. Current approaches rely almost exclusively on dedicated AEA platforms, such as the EA-18G Growler and EC-130H Compass Call aircraft, with high-power, wideband jamming systems. These low density, high-demand systems, however, are intended to primarily operate outside the range of threat systems, so with the increasing reach of the threat, the effectiveness of these stand-off platforms high-power jamming is significantly reduced. At the same time, the beamwidth of their jamming signals is simultaneously expanded with range, making they themselves more visible and vulnerable to a broader range of threats. Improved systems, such as the next-generation jammer (NGJ) being developed for the EA-18G, will somewhat address the problem, but they will not solve it. As pointed out by Bryan Clark, Senior Fellow at the Center for Strategic and Budgetary Assessments (CSBA) (Washington, DC), and co-author of the authoritative study, Winning The Airwaves: Regaining America s Dominance in The Electromagnetic Spectrum, With longer-range air defense systems, we won t be able to generate the same level of effects (with standoff jamming) that we ve been able to in the past. Still, says Clark, although some people have been writing off these big jammers positioned outside the range of enemy aircraft and antiaircraft fire, as sort of a legacy mission, I would argue that there is still a role for stand-off jamming, but it will be much more of a discreet, verynarrow-beam, very-controllable-power level approach, trying to minimize your probability of detection. In any case, additional AEA capabilities will be needed to complement stand-off capabilities. These are expected to take the form of small, low-probability of intercept/low-probability of detection (LPI/LPD) UAVs or expendable systems, such as the Air Force s Miniature Air-Launched Decoy (MALD)-X and the US Navy s version, MALD-N. Clark says, We re already seeing AEA evolving from basically a stand-off, side-lobe jamming approach to something more of a stand-in nature for highly-contested environments. These systems allow you to get in closer and generate effects at lower power without being exposed to attack themselves. AEA AND THE THIRD OFFSET STRATEGY Sitting atop the expectations and plans for next-generation AEA is the

3 e Electronic Attack Approach DOD s new Third Offset Strategy, put forth by Defense Secretary Ashton Carter and Deputy Secretary of Defense, Bob Work. Intended to assure that US forces will continue to have a decisive military advantage over all possible adversaries, a central element of the strategy is the rapid development and deployment of the very latest technological capabilities across all DOD Services and mission areas. Included in the specific areas of technology identified for maximum attention are Deep-learning Systems, Big Data, and Human-Machine Collaboration, all based on the premise that the tight integration of AI and human intelligence will be essential to ensuring this permanent advantage. The US Air Force s EC-130H Compass Call fleet is the mainstay of the Service s stand-off AEA capability. It will soon be replaced by smaller aircraft. (US Air Force) As described by Josh Niedzwiecki, Director of Sensor Processing and Exploitation, BAE Systems (Nashua, NH), The whole idea is to make decisions, and close the decision loop, faster than our adversary. There s a huge growth in AI in commercial industry, and there s now also a lot of interest within the DOD in applying AI and machine learning to our systems to make them smarter and more adaptive in real time and in mission time. Specific to AEA, Niedzwiecki points to the proliferation of wireless telecommunications throughout the electromagnetic spectrum (EMS), which is creating a ton of electromagnetic congestion in the environment, so lots of different and overlapping signals that make it a much more difficult problem to just sort out the signals in the environment to know what to target, what to jam and when to jam. The other dimension is that today s systems are much more software reprogrammable. Whereas in the past, radars were built largely of analog components that took years to field and stayed out there for a long period of time, today, with the proliferation of software-programmable and digital-processing architectures, these threat systems are much more agile making the historical approach of collecting threat data and taking it back to the lab to develop an antidote or prescription for dealing with a new threat-signal type completely impractical. That timeline is way too long and the Third Offset Strategy recognizes this, and that machine learning and advanced signal processing is the way to overcome it. Rick Yuse, President, Raytheon Space and Airborne Systems (McKinney, TX) 39

4 40 sees the same picture. Threat emitters are increasingly agile in the spectral domain with waveforms, frequency, and behavior able to be altered on timescales ranging from pulse to an engagement. This challenge is compounded by commensurate growth in spectral signal density. Agile radars are showing up in unexpected frequencies with waveforms that are difficult to find, particularly when they are obscured in high-signal-density environments. As a result, systems will have precious few threat behavior observations to use before deciding what to do. Learning about a new radar must happen in real time ahead of the threat finding its target. COGNITIVE EW NEEDED TO ADDRESS THE THREAT To meet and keep pace with the threat, Yuse says cognitive systems that can sense the RF environment and adapt to it are needed. Our systems must rapidly respond to dynamic changes in adversary EMS parameters. Cognitive EW, real-time-decision-making and associated tactical execution are priorities. For EMS operations (EMSO), machine learning provides the ability to observe, analyze and act against threats that were unknown before the mission. Pointing to DARPA s extensive ongoing efforts in the area, CSBA s Clark outlines the current state of the technology and the path he expects to see followed as advancements are made. Today our systems are pretty automated, able to generate pre-planned responses to stimuli, but we re now moving into the realm of making these systems adaptable, meaning they can take information and, based on what that information is telling them, be able to take several preplanned responses and come up with a technique or combination of techniques that would be effective. The systems are smarter, but their response is still based on operator input. Clark describes the next step as the place where actual transition to machine learning or AI will take place. We re looking toward EW systems that will be able to detect an unknown signal that doesn t look like anything they ve seen before, but based on its underlying characteristics, be able to figure out a technique that might be effective, use that technique it creates, observe what the effect is, and then change that technique in real time. The technique will continue to change in response to what the machine is seeing in the environment. Clark says this kind of experiential learning is being worked on now at DARPA as part of the agency s efforts in cognitive EW, and that they re currently building demonstrations of the capability. According to BAE s Niedzwiecki, work is, in fact, going on in cognitive EW across the R&D community. We re heavily involved in that process, particularly in machine-learning algorithms. He describes the goal as follows. Instead of looking up and seeing how a new signal matches to things I ve seen before, I use its general characteristics to infer what its capability might be and then I use a series of techniques to try to attack that threat and observe how its behavior changes in response to determine if I m effective. I then use that knowledge to train my algorithms.

5 42 The Miniature Air-Launched Decoy-Jammer (MALD-J) is capable of flying into defended air space and disrupting an IADS from close range. (US Air Force) As Niedzwiecki emphasizes, however, it s not just a potential threat s signal parameters and characteristics that will be involved in the decision-making process, but rather its behavior in general. For example, a weather radar or commercial air-traffic-control radar will have certain behavioral characteristics in addition to signal characteristics, such as scanning the entire air space. These behavioral characteristics will be different for a tracking radar that is targeting a particular aircraft and the signal it emits when it s doing that. These algorithms use that knowledge to help determine which signals get priority and what the best set of techniques are to use against them. Niedzwiecki points out that the EW community is leveraging a lot of the advances in machine learning and in deep learning that the commercial industry has focused on in the big-data domain. We re applying those kinds of things to the EW domain, combining machine learning, advanced signal processing as well as our understanding of physics to solve the problem. In fact, Moore s law (rate at which processing capabilities advance) definitely comes into play. What is nice now is that a lot of these algorithms we re defining can run on very low SWaP processors, so when you think of the machine learning that some of the commercial applications implement, those algorithms live on racks of computers that sit in big network server rooms, but, we re talking about applying those kinds of algorithms to run on embedded systems on a tactical platform. There s a lot of development work being done on optimizing those algorithms to reduce SWaP with minimal effect on performance, as well as on improving the processors, for example through the use of graphical processing units (GPUs) that allow you to do massively parallel computations, and adopting those kinds of processing architectures for EW applications. (For more on cognitive EW, see related story on page 56 of this issue.) NETWORKED OPERATIONS ARE CRUCIAL The ultimate solution, however, doesn t end at the individual platform

6 44 level. It will require the resources of as many different platforms and sensors that can be simultaneously be brought to bear on the problem. And, if the benefits of sensor-/platform-level machine learning are to be truly maximized, each system must be able to instantaneously share all of its learned information with every other system on the network. As referenced by CSBA s Clark, one of the areas emphasized in the Third Offset Strategy is the ability to conduct networked operations. This is part and parcel of the task. EW is electronic protection (EP), electronic support (ES) and electronic attack (EA), and you re not going to be able to do networked warfare unless you can manage and control the EMS where the networks operate and live. I would offer that our advantage in networked warfare is at risk because of that exact problem. We may upgrade networking technology and have lots of experience with it, but because it depends on our access to, and ability to use, the EMS, if our adversary goes after our access to the EMS, all that networking technology will be for not. That is why there is so much attention being paid to EW in the Third Offset to ensure that we have the access to the EMS required to facilitate that networked capability. Raytheon s Yuse echoes this point. The power of the simultaneous convergence of technology, which will integrate and optimize employment of capabilities across mission systems and ultimately the force package, is at the heart of the strategy. Strategically designing and deploying a successful AEA capability requires systems that adapt to previously unknown behavior, while preserving legacy effectiveness. These systems must integrate multiple sensing modalities and coordinate multiple effects. We re investing in transforming EMSO from strictly a mission enabler to an effector in its own right. A variety of integrated technologies and systems will have a part to play in the technology and weapons mix. High-power standoff jammers, small light-weight AEA systems on non-traditional platforms, highly-integrated, complex systems on high-value platforms will be included. It will take layers of defense, ranging from kinetic and non-kinetic EMSO effects, to defeat threats. To reach the ultimate capability currently envisioned for AEA, major advancements in networking technology will be necessary. Says Niedzwiecki, Networking is always being asked to push further because of a few different factors. One is that the EMS is becoming more and more crowded and so finding available spectrum to communicate without being interfered with is an increasing challenge. This, of course, is in addition to adversaries actively attempting to jam your communications signals. So, operating and providing networking for these systems in an A2/AD environment is a challenge, and requires that we apply the same concepts being advanced for cognitive AEA to cognitive EP. How do I leverage my understanding of the jamming environment on my communication signals to optimize in real time my waveform and maintain communications in the presence of jammers.

7 46 Another dimension of the AEA networking challenge relates to network capacity and data rate, with an ever increasing need to pass more and more information around the network. To accomplish this effectively requires highly efficient network-optimization techniques. Niedzwiecki says this is indeed another area that is receiving significant attention, making sure that network resource management and network coordination algorithms are resilient enough to handle dropped data or Any platform. Any time. db Control delivers. TWT AMPLIFIERS MICROWAVE POWER MODULES POWER SUPPLIES Trust db Control with your high-power solutions for radar, ECM and EW threat simulation systems. Our TWT Amplifiers (TWTAs), microwave power modules (MPMs) and power supplies can be custom-configured for missioncritical ground-based, shipborne and highaltitude manned/unmanned airborne platforms. We utilize modular designs for rapid configuration to create power products that are reliable and easy to maintain. If your expectations are as high as your power requirements, contact us today. missing information. This is especially important where the signal density in the environment from an EW perspective is high the ability to manage your resources and have, for example, one platform targeting certain signals and another platform targeting others. +1 (510) db Control Corp. All rights reserved. Photos provided courtesy U.S. Air Force, U.S. Army and U.S. Navy. MERGING OF COMMS AND RADAR AEA In addition to the evolving threat, advances in machine-learning algorithms, processing, and networking technology will also impact the types, mix and missions of platforms performing AEA. On the mission side, communications and radar AEA missions, which are currently largely served by different, specialized platforms and systems will be increasingly met with single, multi-purpose systems. Says CSBA s Clark, In the past, we ve had to treat them separately because when you re using hardware-based signal processing and jamming systems (TWTs, waveguides etc.) you don t have that adaptability. The hardware drives the frequency range and the characteristics of the signals that you can generate. Given that threat radars and communication systems have generally operated in different frequency ranges, this has meant that different AEA systems and the platforms carrying them have also been different. Now, Clark says, however, with the advent of GaNbased AESA technology, we can expect to see more multi-function, softwarecontrolled systems able to operate over much wider bandwidths performing both missions. We re already seeing some of this today where, for example, some radar-jamming systems have the ability to go down into the comms frequency ranges and vice versa. We ll see more of this with NGJ and with SEWIP Block 3 in the Navy, and it s inevitable that we ll see common-purpose AEA systems for both comms and radar. Raytheon s Yuse agrees but adds that, AEA capabilities will need to be incorporated into sophisticated, scalable, affordable end-to-end EMSO systems, and the future of intelligent EMSO is all about networks, high-gain, EA, cyber, scalable common back ends and multi-function arrays. A mix may be effective, but will require networked operation that comes with the inherent challenges of distributed management and communication. Multi-purpose systems also face challenges, however. Multi-purpose tools are compromised to get something effective for multiple uses, but suboptimal for one use. Systems with primary and secondary functions (or redundancy) may be a good compromise and contribute to overall combat resiliency. BAE s Niedzwiecki says this leads back to the subject of resource manage-

8 48 DARPA s Gremlins program seeks to prove that swarming, networked unmanned platorms can carry out a variety of missions, including AEA, inside an IADS. (DARPA) GaN and GaAs Solid-State Power Amplifiers for Multi-Function, Radar, and EW System Design Whether your application is narrowband, wideband or ultra-wideband, operating in pulsed or CW mode, CTT s power amplifiers are an especially attractive choice for new multi-function frequency-agile systems that effectively conserve weight, space and power consumption. The characteristics of the portion of the electromagnetic spectrum selected for any of these particular system designs are undoubtably the most important to the end user, as it has the greatest impact on the type of information required and received. Engineered specifically to meet the stringent requirements imposed by many modern system designs, CTT s family of GaN and GaAs-based solid-state power amplifiers excel in a wide range of applications. CTT has delivered production quantities of amplifiers with power levels from 10 through 200 Watts and higher for a variety of multi-function, radar and EW applications. AMDR Shipboard Radar AESA Radar VLO/FLO Threats New Land Radar EW UAVs More than thirty years ago CTT, Inc. made a strong commitment to serve the defense electronics market with a simple goal: quality, performance, reliability, service and on-time delivery of our products. Give us a call to find out how our commitment can support your success. It s that simple. New Catalog Available! Enabling wideband frequency agility Microwave Technology Leadership Power Amplifiers NEW GaN and GaAs Models Radar Bands up to 1kW EW Bands up to 400W Pulse and CW Solid-State Microwave Power Modules Rack-Mount Configurations Low-Noise Amplifiers Up and Downconverters Subsystems Custom Engineered Options USA-based thin-film microwave production facility ISO IS9100C 9001:2008 ISO QUALITY 9001: East Java Drive Sunnyvale California Phone: Fax: sales@cttinc.com ment. When you re looking at multiple signals in an environment, the algorithms will have to figure out which of those signals are the most important, which are the most lethal, and which are the highest priority for me to target with my system to maximize overall mission effectiveness. This, in turn, relates to the mix of platforms or systems that will best meet the overall AEA requirement and the need for effective resource management at the EW level. This can then also be extrapolated to the broader-level, mission-management task, and the extent to which systems will operate autonomously. With access to all this information, observes Niedzwiecki, mission commanders will have a better picture of the threat space so that they can make those tactical decisions during a mission. On the other hand, lessening the human taskload is one of the most promising benefits of the application of cognitive EW and machine learning to the AEA mission. To describe the alternatives and tradeoffs to be considered, Niedzwiecki points out that there are certainly configurations and mission sets where you would want these systems to be completely autonomous because of the timelines required, but there are also others where you may want to have an operator on-the-loop. In this scenario, while an operator is not actually in-the-loop driving the decision process, they can observe what is going on and confirm an approach that the machine recommends, or in certain cases, manually override it. The third option is the operator inthe-loop, which Niedzwiecki describes as more of a mission commander, where you re not only using the information you collect and process about the EMS, but you re also using all the information that you re collecting from all the other sensors on your platforms to help optimize your overall mission. The mission commander is where some of that human-in-the-loop can take over. Even so, Niedzwiecki notes that there is also ongoing research looking at automating that process as much as possible, as well, in order to reduce the cognitive load of the warfighter.

9 50 BEYOND COMMS AND RADAR-BASED THREATS Radars and communication systems will not be the only threat systems contributing to the challenges of penetrating a sophisticated IADS or A2/AD environment. Other potential surveillance, targeting, and weapons technologies, both active and passive, will also have to be considered and addressed such as the use of commercial emitters (television and radio signals) cueing passive radars, satellite-based systems, IR and other optical sensors, UAVs, and man-portable air defense systems. CSBA s Clark says he believes the DOD recognizes the need to include these potential threats in its AEA mix and capabilities, but also sees them being constrained by the fact that the Services haven t yet generated operational concepts that will drive requirements for them. That s one of the problems we have. Even though the technology community is pursuing lots of options that could improve our ability to detect and engage these different targets, they don t have requirements to do that. The requirements are based on what the Services/warfighters are thinking they will get into. So there is a little chicken-andegg thing going on where the technologists need to portray to the warfighters what the level of technology is going to be and what could be possible, and the warfighters need to start incorporating those into requirements so they can build the programs based on them. There needs to be that OODA loop (decision cycle of observe, orient, decide, and act) of technology that can offer certain potentials, and then the warfighters can figure out how to use those potentials and turn them into requirements. That cycle needs to be sped up or else you just never get the acquisition process to move at the speed that the technology can provide. Although Raytheon s Yuse says that ultimately systems will in fact address other surveillance and targeting threats, he adds that, it may not be the AEA device (alone) that will do it. A directed energy (DE) effect, onboard laser, may blind the sensor, while a jammer simultaneously engages the target, increasing overall mission effectiveness. A flexible, digital receiver exciter can be an effective weapon against a command, control and communication infrastructure, denying detection as sure as jamming the radar itself. It s not a stretch to think about a jammer that could very well be effective against a wide array of systems other than radars. The ability to attack the integrated air defense system rather than single radars is what we need. SERVICES ROLES AND THE SYSTEM OF SYSTEMS (SOS) Although the DOD and the Third Offset Strategy is decisively supportive of the smart-system approach to future AEA, it has not been specific about the particular operating concepts that will be used, or the roles of the individual Services, leaving this largely to the Services themselves to determine. CSBA s Clark, says he doesn t see this changing, at least in the near term. Rather, he believes the DOD sees its role more oriented toward getting the Services to pursue

10 52 new capabilities in particular areas and then transitioning out of others. Their intent is to accelerate the innovation they need in EW without necessarily providing direction, like differentiating AEA roles and missions by Service. There will still be a kind of pick-up game there with regard to who does what aspect of the AEA mission based on what the individual Services own operating concepts are. In the end, however, future AEA will certainly represent a combination of sensors, jammers, and actual weapons platforms that will be capable of providing the required effects in other words, a system of systems (SOS) approach. And, in fact, included in the 2017 Air Force RDT&E budget is an activity aimed at providing just that. Titled AEA System Engineering Studies & Technology Transition, the line item represents $9.2 million for overall systems engineering, modeling and simulation, architecture and network requirements development, effectiveness assessment and requirements allocation to component systems of the Airborne Electronic Attack (AEA) System of Systems (SOS). The program will also assess operational effectiveness of multiple EW systems in both offensive and defensive roles. The program description also notes that the program is included in Budget Activity 5, System Development and Demonstration, because it has passed Milestone B approval and is conducting engineering and manufacturing development tasks aimed at meeting validated requirements prior to full-rate production. Target accomplishments for FY 2017 call for support of the joint Air Force/Navy offensive AEA SOS Analysis of Alternatives, update of the AF EW Roadmap, and a report on preferred concepts that provide advanced cost and operationally effective materiel solutions to joint AF/Navy combat operations in the time frame. CSBA s Clark, says the effort is part of the DOD s work to rationalize their investments in AEA between the Air Force and the Navy. The Services won t be told how to divvy up the mission, but this SOS work will inform ways that they can be more efficient, because each Service is trying to find ways to spend less money, and to avoid duplication of effort. Even so, he expects there will be some effort to try to transition the program s eventual recommendations into some form of roles-and-missions differentiation, as well as delineation of the new operating concepts that the Services are expected to be pursuing. The work will be done within each of the Services, led from within the AF and Navy headquarters, but using the technical expertise of the System Commands, says Clark. In summing up his view of the current state of AEA affairs, Raytheon s Yuse says that while his company is actively working to develop solutions for these challenges, transitioning these new capabilities to users fast enough is just beginning to be addressed. Ensuring users trust the capability enough to fully and effectively implement is key, and spectrum dominance success will ultimately come from Service collaboration to achieve Joint Force Commanders objectives and making EMS dominance a National security priority. a