Target scene frequency diversity exploitation with ultra-wideband orthogonal frequency division multiplexing radar

This paper concerns experimental investigation of the frequency-domain approach to target scene sensing using ultra-wideband orthogonal frequency division multiplexing software-defined radar. Individual spectral components of the radar signal (sub-carriers) are analysed with the aim of performing their adaptive selection and establishing similarity of the observed response to the previously collected responses of various target scenes. By performing this analysis, the authors are effectively utilising frequency diversity of targets for detection purposes. A low-power, short-range radar system built at Miami University is used to apply this approach to target detection with generalised-likelihood ratio test (GLRT) and frequency profile matching (FPM). Experimental results indicate good promise of FPM, which performed equally to, or outperformed GLRT in all scenarios. Results of correct detection and false alarm rates are shown and discussed, along with theoretical descriptions of the detection approaches.