Blind channel estimation for fading channels with Markov inputs
C. Carlemalm, A. Logothetis · 2002
In this paper, we address the problem of estimating the input sequence generated by a finite state Markov source transmitted through a wireless communication channel with unknown parameters. The channel suffers from fading and scattering effects and is modeled by a finite impulse response (FIR) filter with stochastically time-varying taps. The aim is to jointly estimate the transmitted sequence and the state of the time-varying channel. The proposed iterative scheme optimally interconnects a hidden Markov model (HMM) smoother and a Kalman smoother to yield state estimates, while Yule-Walker and Baum-Welch type equations yield the channel FIR and the Markov chain parameter estimates, respectively.