Adaptive array beamforming for cyclostationary signals

This paper concerns the adaptive array beamforming using signal cyclostationarity. We present recursive algorithms for implementing the self-cohesive restoral (SCORE) approaches which were originally developed by Agee, Schell and Gardner (1990) to perform blind adaptive signal extraction using antenna arrays. Utilizing the theory of matrix factorization in conjunction with a power normalization scheme, we first reformulate the original problems considered by the SCORE approaches as simpler constrained optimization problems. By employing the modular Gram-Schmidt orthogonalization (GSO) structure, adaptive algorithms are then developed for finding the optimal weight vector. It is shown that although the GSO structure used requires a computation load O(N² ) for an adaptive-beamformer with N elements, the computing time required for producing an output data vector is reduced to O(N) due to the pipeline implementation suitable for the GSO process. The convergence property for each of the proposed algorithms is considered. A simple and efficient method is proposed for choosing appropriate initial guesses to initiate the adaptive algorithms. Simulation examples are given to show the effectiveness of the proposed algorithms