Iterative joint decoding and sparse channel estimation for single-carrier modulation

Single carrier (SC) modulation over a sparse channel motivated by underwater acoustic communications (UAC) is considered. A BER analysis for deterministic channels is first presented that identifies conditions under which SC with zero forcing detection outperforms OFDM. The analysis also motivates the use of MMSE detection in Rayleigh fading channels. An iterative receiver is then developed combining LDPC decoding, channel estimation under a numerosity constraint and CFO tracking. It is shown that MMSE turbo equalization arises naturally from the marginalization in the factor graph when the decoder extrinsic densities are approximated as Gaussian. A frequency-domain MMSE turbo equalizer is then combined with matching pursuits (MP) channel estimation. Simulation results for a realistic UAC channel are given showing consistent improvement using iterative processing.