Multistatic human micro-Doppler classification with degraded/jammed radar data

This paper investigates the classification performance when using multistatic human micro-Doppler radar data that have been degraded by some form of jamming. Two simple cases of Signal-to-Noise Ratio (SNR) degradation and nulling of a sub-set of the available radar pulses are considered for these initial results, leaving more complex forms of jamming or degradation for future work. Experimental data collected with a multistatic radar are used in this study, aiming to classify 7 similar human activities, when individual subjects are walking carrying different objects. The results show that the use of multistatic radar data can provide resilience to the effect of the data degradation, thanks to the redundancy and additional information available from multiple radar nodes.