Air Surveillance Radar Classification Improvement

Increasingly capable Unmanned Aerial Systems (UASs) have made the accurate identification of such threats an important aspect of battlefield situational awareness. Army Counter-fire Target Acquisition (CTA) radars, including the man-portable AN/TPQ-50 and larger long-range AN/TPQ-53, provide the greatest opportunity to incorporate more advanced air target classification techniques because they are widely fielded. In Phase I, Technology Service Corporation (TSC) will focus on state-of-the-art Machine Learning algorithms, including Support Vector Machines (SVMs) and Neural Networks (NNs), to accurately discriminate UASs, from manned aircraft, birds, ground vehicles, and clutter. SVMs and NNs, which can also provide a classification confidence level, have proven successful for target classification in VADER and will be explored first. Minor changes to the CTA radar waveforms, ensuring that the primary surveillance mission is not impacted, will also be considered to optimize UAS identification. To ensure a successful transition of this technology, and to obtain ample data from actual CTA radars, TSC is teamed with SRC. As prime contractor on the TPQ-50, SRC can provide expertise on this sensor’s capabilities and useful information on variants including LSTAR that were developed to address these threats. In Phase II, TSC will mature the most promising approaches and conduct field tests.