Frequency Tracking Performance of Adaptive Lattice Filters

摘要

Adaptive linear prediction filters have been proposed for the detection ofnarrowband signals in nonstationary noise and for the removal of narrowband interference from spread spectrum communications signals. In this paper, analytical results and computer models for the PARCOR coefficients, signal recovery error, and output misadjustment of the stochastic gradient adaptive lattice filter will be obtained for a complex linear chirp FM signal in white, Gaussian noise. From these expressions an optimal adaptation rate is obtained as a function of the filter order and the chirp rate. These results illustrate that short filter lengths are sufficient for tracking chirp signals. submarine systems, adaptive signal processing, adaptive line enhancing, adaptive noise cancelling.