A new class of least-squares algorithms is pre- sented foradaptive filtering. The idea is to use a foxed set of directions andperform line search with one direction at a time in a cyclic fashion.These algorithms are called Euclidean Direction Search(EDS)algorithms. The fast version of this class is called TheFast-EDS or FEDs algorithm. It is shown to have O(N) Computationalcomplexity and a convergence rate comparable to That of the RLSalgorithm. Computer simulations are presented To illustrate theperformance of the new algorithm.
展开▼