Knowledge-aided conjugate gradient algorithm for adaptive space-time processing in airborne radar

The convergence nature of conventional conjugate gradient (CG) algorithm often result in degraded system performance. This paper presents a new knowledge-aided CG (KACG) approach to the cancellation of clutter echoes in modern AEW radar systems equipped with digital beamforming. KACG use finite precision knowledge of the radar system parameters and flight geometry knowledge to perform initialization. Results of experiments indicate that the improvement in convergence can be substantial. Compared to the original CG approach, KACG can be regarded as practical solutions for robust and effective space-time adaptive processing (STAP).