Blind testing for equalisation errors for time-varying communication channels

This paper presents algorithms for blind detection of equalisation errors when the communication channel is a finite impulse response (FIR) system with stochastic time-varying taps. A linear autoregressive (AR) model is used to describe the FIR tap behaviour. Error detection is achieved by a binary hypothesis test on the innovations obtained from a Kalman filter. We present off-line and on-line error detection algorithms. These algorithms are experimentally verified in computer simulations.