Improved LMS estimation via structural detection

摘要

In various adaptive estimation applications, the unknown channelis characterized by an impulse response which consists of extendedregions of negligible response or `inactivity'. We consider the LMSestimation of a channel that may be well approximated by an FIR (filter)model with only a few nonzero tap coefficients within a given delayhorizon or tap length n. When the number of nonzero tap coefficients mis small compared to the delay horizon n, the performance of the LMSestimator is greatly enhanced when this specific structure is exploited.We propose a consistent algorithm that performs identification ofnonzero taps only