A generalized MIMO architecture for set-membership-based signal processing

A generalized multiple-input-multiple-output (MIMO) parallel architecture is developed. It implements the complex-vector case of the set-membership weighted recursive least squares (SM-WRLS) algorithm, and very significantly improves the efficiency of this algorithm when compared to the conventional matrix inversion lemma-based WRLS algorithm. This compact architecture uses only &ogr;(m) cells and reduces the computational complexity of the algorithm to &ogr;(m ) operations per data set, where m represents the number of parameters to be estimated in each dimension of the model