A hidden Markov model-RPE algorithm for narrowband interference suppression in spread spectrum systems

Abstract

This paper presents a novel nonlinear estimation algorithm based on hidden Markov models for narrowband interference suppression in CDMA spread spectrum systems. The proposed algorithm combines a recursive hidden Markov model estimator, a Kalman filter and a recursive prediction error parameter estimation algorithm. It is shown that the proposed algorithm not only outperforms the nonlinear filtering techniques for narrowband interference suppression presented in Rusch and Poor (1994), but that it outperforms and has faster, more robust convergence properties than the cross-coupled expectation maximization based algorithm presented in Logothetis and Krishnamurthy. © 1998 IEEE.