Detection of behind-the-wall targets with adaptive UWB OFDM radar: Experimental approach

In this paper the problem of target detection with an adaptive radar in an indoor environment is considered from an experimental perspective. Statistical models of scenarios with and without targets were generated from collected data using an ultra-wideband (UWB) orthogonal frequency division multiplexing (OFDM) software-defined radar system. This system allows for independent processing of individual frequency components (sub-carriers). The responses in each scenario for each sub-carrier were used to construct distribution models, as well as to perform maximum likelihood estimations (MLE) of distribution parameters and adaptive detection thresholds for each sub-carrier were derived for constant false alarm rate (CFAR) performance. The Generalized Likelihood Ratio Test (GLRT) was then performed in two experimental scenarios and conclusions about detection and adaptive pulse design using OFDM sub-carrier weights were drawn.