Multi-channel Recursive Least-Squares Algorithm Adaptive Filter

• Main Authors: Hsing-Hsuan Chien, 簡杏軒
• Other Authors: Te-Jen Su
• Format: Others
• Language: zh-TW
• Published: 2006
• Online Access: http://ndltd.ncl.edu.tw/handle/82854445515589945692

碩士 === 國立高雄應用科技大學 === 電子與資訊工程研究所碩士班 === 94 === The well-known code division multiple access (CDMA) decorrelating detector (DD) uses a bank of correlators, followed by the inverse of the matrix operation to eliminate the multiple access interference (MAI). However, the operation for the inverse of the matrix involves a great deal of computation, especially when the users’ number is large. Therefore, in this thesis, we propose some recursive methods, the Least-Mean-Square (LMS) algorithm and the Recursive Least-Squares (RLS) algorithm, to detect users’ signals adaptively. We make use of the analogy between a traditional Winner filter and the decorrelating detector to construct adaptive implementation schemes of the decorrelating detector, called decorrelating transversal filter . We applied both LMS algorithm and RLS algorithm to the decorrelating transversal filter, just as the ways to apply the LMS algorithm and RLS algorithm to the Winner filters. With the proposed schemes, we can greatly reduce the computational complexity of a CDMA multi-user detector while maintaining an acceptable performance.