Stochastic resolution analysis via a GLR test in radar

Abstract

Stochastic description of resolution is explored that includes probability of resolution and signal-to-noise ratio (SNR). Taking SNR into account is especially relevant in compressive acquisition typical for compressive sensing (CS) due to the fact that fewer measurements are acquired. Our stochastic approach to resolution uses information distances computed from the geometry of data models that is characterized by the Fisher information. The probability of resolution is assessed via a test with the generalized likelihood ratio (GLR) by exploiting information distances in asymptotic GLR. Based on this information-geometry approach, we demonstrate the stochastic resolution bound in test cases with spatial and temporal measurements for joint angle-Doppler processing. In addition, we also compare such stochastic resolution bounds with actual resolution obtained from sparse-signal processing that is nowadays a major part of the back end of a radar system with CS. Numerical results demonstrate the suitability of the proposed stochastic resolution bounds in (radar-)resolution analysis because of its sensitivity to crucial impacts on the performance guarantees: array configuration as well as input SNR, separation and a probability of resolution.