Soft Input Iterative Channel Estimation for Turbo Equalization over Time Varying Frequency Selective Fading Channels

Availability of channel state information (CSI) at any communication receiver is important for accurate detection and decoding of transmitted signals. Different types of algorithms have been proposed for the estimation of channel impulse response (CIR) of a time varying channel among which are the least mean square (LMS) and the recursive least square (RLS) algorithms. In this paper we propose a variable step size algorithm for adaptation of CIR estimate by channel estimator incorporated into turbo receiver for turbo code. The soft information feedback from the turbo decoder, which is based on low-complexity soft-input/soft-output (SISO) equalization algorithm of a time-varying CIR, is used to improve the channel estimate over a number of iterations. The channel estimate is employed by a turbo equalizer to remove the effect of intersymbol interference imposed by the communication channel. The proposed algorithm alongside other algorithms has been used for channel estimation in combination with the equalizer. The simulation results for normalized fade rates 0.01 D s f T = and 0.005 D s f T = show how the proposed channel estimation basedalgorithm outperforms LMS-based channel estimation algorithm and its performance is very close to that of RLS-based channel estimation algorithm. Keywords– Iterative channel estimation, soft input, turbo code turbo decoding, variable step size NLMS algorithm.