The computation complexity of RLS algorithm is on the order of O(K2) per sample, which is a challenge for implementation, especially when K is large. For comparison, the computation complexity of LMS algorithm is on the order of O(K) per sample. In general, the convergence speed of RLS algorithm is an order faster than LMS algorithm. In a word, all the three conventional algorithms listed above cannot achieve fast convergence speed and little computational complexity at the same time.