GRADIENT BASED SUBSPACE TRACKING ALGORITHMS AND SYSTOLIC IMPLEMENTATION
Bin Yang · International Journal of High Speed Electronics and Systems · 1993
A variety of modern signal processing applications like high resolution temporal/spatial spectral analysis and eigenvalue decomposition based data compression requires subspace estimation. In an unknown and possibly changing environment, adaptive algorithms which are computationally efficient, numerically stable, and easy to implement in hardware are highly desirable. In this paper, we first review some gradient-based adaptive algorithms for subspace tracking. We show that these algorithms, in comparison to other classical techniques, are not only competitive in tracking performance but also advantageous in both computational complexity and robustness. Then we present a novel systolic architecture for implementing these algorithms with a high processor utilization.