Boosting multi‐target recognition performance with multi‐input multi‐output radar‐based angular subspace projection and multi‐view deep neural network

Current radio frequency (RF) classification techniques assume only one target in the field of view. Multi-target recognition is challenging because conventional radar signal processing results superposition micro-Doppler signatures, making it difficult to recognise multi-target activity. This study proposes an angular subspace projection technique that generates multiple data cubes (RDC) conditioned on angle (RDC-ω). approach enables separation raw RDC, possible utilisation deep neural networks taking RF as input or any other representation scenarios. When targets are closer proximity and cannot be separated by classical techniques, proposed boosts relative signal-to-noise ratio between targets, resulting multi-view spectrograms accuracy when DNN. Our qualitatively quantitatively characterise similarity signatures those acquired a single-target configuration. For nine-class activity problem, 97.8% 3-person scenario achieved, while utilising DNN trained data. We also present for two cases close (sign language side-by-side activities), where has boosted performance.