An iterative receiver for time-domain oversampled OFDM system over time-varying underwater acoustic channels

Abstract

This paper investigates an iterative receiver for time-domain oversampled orthogonal frequency division multiplexing (OFDM) over time-varying underwater acoustic channels for both single-input, single-output (SISO), single-input, multiple-output (SIMO) and multiple-input, multiple-output (MIMO) scenarios. We extend two widely used parametric models for time-varying channels: (i) Doppler rate model (DRM) and (ii) basis expansion model (BEM) to work with oversampled OFDM systems. To exploit the inherent sparsity of underwater acoustic channel, iterative orthogonal matching pursuit (OMP) algorithm is adopted for channel estimation. To further reduce computational complexity, fast Fourier transform (FFT) based low-complexity implementations for iterative OMP algorithm are proposed for both BEM and DRM. The proposed receiver is evaluated by sea trial conducted in the Trondheim Fjord, Norway. Both 8-phase shift keying modulation scheme and 16-quadrature modulation (16-QAM) scheme are carried out during the sea trial. The experimental results show that the proposed iterative time-domain oversampled OFDM receiver achieves lower bit error rates (BERs) than conventional iterative symbol rate sampled OFDM receiver.