An Adaptive Hybrid GLRT-Bayesian Detector for MIMO Radar in Non-Stationary Clutter Environments

Abstract

This paper proposes a novel adaptive hybrid detector for MIMO radar operating in non-stationary clutter environments. Classical Generalized Likelihood Ratio Test (GLRT) detectors require a large number of training samples, while Bayesian detectors rely on fixed prior information that becomes obsolete when the clutter statistics change over time. To overcome these limitations, we introduce a sliding-window mechanism that dynamically updates the Inverse Wishart prior in real-time. The proposed detector fuses the GLRT and Bayesian statistics into a single hybrid metric. Simulation results demonstrate that the proposed detector significantly outperforms both classical GLRT and fixed-prior Bayesian detectors under dynamic clutter scenarios, maintaining high detection probability with low computational complexity.