The authors present the use of an LMS (least mean square)-based algorithm called the noncanonical LMS (NCLMS) for channel equalization. The NCLMS algorithm is not implemented on a standard FIR (finite impulse response) filter, but uses a structure referred to as the noncanonical FIR (NCFIR). When the NCLMS is implemented the signal flow graph can be shown to be fully pipelinable and the sampling rate is not inhibited by the recursive nature of the algorithm, as is the case for the LMS. Channel equalization results are given which show that the NCLMS has a reduced excess mean squared error level and an improved performance in an impulsive noise environment over the LMS.
展开▼
