A systolic array for recursive least squares computations
Marc Moonen, Joos P. L. Vandewalle · IEEE Transactions on Signal Processing · 1993
A triangular processor array has been developed for implementing recursive-least-squares computations. The throughput is O(n/sup 0/), which means that data vectors are fed in at a rate that is independent of the problem size. The corresponding least-squares solutions then run out of the array at the same rate. The overall efficiency is roughly 67%. This array is believed to be the first to implement a complete RLS algorithm in a stable and efficient manner.>