Multi-channel Fast Recursive Least Squares Adaptive Filtering Algorithm for Parallel Implementation Using Four Processors

A new regularized fast RLS algorithm with the sliding window is presented for adaptive filtering. It is to be implemented using four processors in parallel. The algorithm is obtained for the general case of multi-channel adaptive filters having unequal numbers of complex-valued weights in channels. This algorithm can be used for adaptive filtering of non-stationary signals and as a component of the regularized sliding window fast RLS algorithms with linear constraint.