点击次数：

论文类型： 会议论文

发表时间： 2004-08-31

收录刊物： EI、CPCI-S、Scopus

卷号： 1

页面范围： 392-395

关键字： adaptive filter; RLS; linear prediction

摘要： In this paper, a new computationally efficient algorithm for recursive least-squares (RLS) algorithm called Reduced Order RLS. it is also called as Partial Updating RLS (PU-RLS), algorithm is introduced. The basic idea of UP-RLS algorithm is that a high order filter function is decomposed into two low order simple functions, and then update the sub filter coefficients partially which result in less computation. This kind of adaptive algorithm shows much more enhanced computational efficiency compared to the earlier works such as Split RLS algorithm but with better performance than Split RLS algorithm based on simulation results.