An improved direct adaptive multichannel turbo equalization scheme for underwater communications

An improved direct adaptive multi-channel turbo equalization receiver structure employing adaptive Doppler compensation is proposed for wireless communications over sparse underwater acoustic (UWA) channels. Compared to existing direct adaptive turbo equalization schemes that have been used for underwater communications, two novel techniques are employed to improve the overall system performance: 1) the extrinsic log likelihood ratio (LLR) output by the equalizer, and exchanged between the turbo equalizer and decoder is calculated by taking into account the a prior probability of the symbol constellation other than assuming equal probability, 2) improved proportionate normalized least-mean-square (IPNLMS) algorithm that can take advantage of the sparsity inherent in underwater acoustic channel is utilized to update the coefficients of the multi-channel turbo equalizer. Lake experiments results show that the proposed scheme can achieve 2kbps error-free data transmission in 1.8km range, 2kHz bandwidth with a carrier frequency of 9kHz, and obviously outperform the existing direct adaptive turbo schemes for underwater communications in convergence rate and bit error rate (BER).