Cue Aggregation Algorithms for Multi-Function Receivers

GIRD Systems, Inc. (GIRD) proposes to invent and develop FOCUS ñ Flexible Optimization of CUe Software. FOCUS will be built on a set of new machine learning (ML) algorithms that run on commercial-off-the-shelf (COTS) processors to both aggregate multiple signals into classes of same pattern of emissions and to locate methods to distinguish look-alike signals from platforms with different spectral signatures or different intentions. FOCUS is motivated by the progress in Analog to Digital Converters clocking at over 10 GHz outpacing signal processing resources so that there is a need to prioritize and schedule signal processing resources for wideband multi-function receivers based on the signals in the environment and guidance from a user / supervisory system as to what types of signals are of interest. ÝÝFOCUS is responsible for 1) generating and processing cues (some of which might be low-level and from a third party), which are metadata about which of several different signal classes are present in a forensic buffer and where (frequency, time, ID), and providing that information to a DSP resource scheduler being developed under a different SBIR solicitation, 2) combining or separating classes of signals that are relevant to the userís desires, 3) training new cue generators (effectively ML signal classifiers), 4) identifying relationships between signals that may be useful for downstream signal processing (e.g., phase shifts across antenna elements or variations in signal distortions).