Simplified Kalman estimation of fading mobile radio channels: high performance at LMS computational load

Abstract

Low-complexity algorithms for channel estimation in Rayleigh fading environments are presented. The channel estimators are presumed to operate in conjunction with a Viterbi decoder, or an equalizer. The algorithms are based on simplified internal modeling of time-invariant channel coefficients and approximation of a Kalman estimator. A novel averaging approach is used to replace the online update of the Riccati equation with a constant matrix. The associated Kalman gain is expressed in an analytical form. Compared with RLS (recursive least squares) tracking, both a significantly lower bit error rate and a much lower computational complexity are achieved.<>