Identifying and Characterizing Cognitive Sensor Systems in Tactical Environments

Adversary radar jamming capability will greatly increase in the near future.  Improvements in radio frequency (RF) hardware, especially in solid state RF electronics, will lead to improved jammer capability with reduced size, weight, and power requirements.  Moreover, improvements in embedded computer systems will give jammers access to powerful machine learning and artificial intelligence algorithms to improve their effectiveness.  will challenge current and future electronic protection systems. A strategy to mitigate the capability of cognitive jammers (e.g., sophisticated, AI-driven jammers that reacts to measured radar emissions and measure their own effectiveness) is to attempt to model and exploit the decision logic used by the jammers to select electronic protection techniques which mitigate the jammer effects.  Vadum will develop the Adversary Logic Exploitation System (ALES) to probe a cognitive jammer system in order to model the jammer decision logic.  ALES will use the modeled decision logic to schedule subsequent radar waveforms which both produce the information required by the radar and limit the effect of electronic attack on the radar system.  The ALES algorithms will analyze incoming signals and segregate out the likely jamming signals.  Changes in radar behavior which drive changes in jammer behavior are saved in a persistent data structure (the Jammer Behavior Object or JBO) and analyzed.  Based on the inferred understanding of the jammer behavior, subsequent radar waveforms which avoid the effects of the likely electronic attack will be scheduled.  ALES will identify and schedule both electronic protection techniques and probing waveforms to not only protect the radar, but to gain additional information about the jammer. Vadum will develop the ALES algorithms in an in-house modeling and simulation environment which supports concurrent experimentation and development of algorithm prototypes and production of software versions of those same algorithms.  The Phase I effort will prove out the ALES concept sufficiently to support additional Phase II research focused on improving ALES performance by further algorithm development and characterization along with the potential inclusion of algorithms and architectures from other Vadum Cognitive Electronic Warfare programs.  The Phase II program will also increase the maturity of the algorithms themselves through increased testing fidelity to support eventual Phase III transition decisions.