Resonance-region radar target identification using aspect sampling

We investigate the performance of three target identification algorithms that rely on aspect samples of target features extracted from multifrequency resonance-region radar returns. The first algorithm uses only the amplitudes of radar returns, the second uses amplitudes in conjunction with feature-space trajectories, and the third uses complex radar returns and time-domain correlations. To test the algorithms, the Numerical Electromagnetic Code was used to generate radar returns of five test targets made of conducting wires. Simulation results show that very low identification error probabilities can be achieved with relatively large sampling intervals, few frequencies and low computing costs.