Underwater wireless optical communication utilizing low-complexity sparse pruned-term-based nonlinear decision-feedback equalization

Abstract

The nonlinearity of the light-emitting diode (LED) in underwater wireless optical communication (UWOC) systems is considered the one major limiting factor that degrades the system’s performance. Volterra series-based nonlinear equalization is widely employed to mitigate such nonlinearity in communication systems. However, the conventional Volterra series-based model is of high complexity, especially for the nonlinearity of higher-order terms or longer memory lengths. In this paper, by pruning away some negligible beating terms and adaptively picking out some of the dominant terms while discarding the trivial ones, we propose and experimentally demonstrate a sparse pruned-term-based nonlinear decision-feedback equalization (SPT-NDFE) scheme for the LED-based UWOC system with an inappreciable performance degradation as compared to systems without the pruning strategy. Meanwhile, by replacing the self/cross beating terms with the terms formed by the absolute operation of a sum of two input samples instead of the product operation terms, a sparse pruned-term-based absolute operation nonlinear decision-feedback equalization (SPT-ANDFE) scheme is also introduced to further reduce complexity. The experimental results show that the SPT-NDFE scheme exhibits comparable performance as compared to the conventional NDFE (nonlinear decision-feedback equalization) scheme with lower complexity (the nonlinear coefficients are reduced by 63.63% as compared to the conventional NDFE scheme). While the SPT-ANDFE scheme yields suboptimal performance with further reduced complexity at the expense of a slight performance degradation, the robustness of the proposed schemes in different turbidity waters is experimentally verified. The proposed channel equalization schemes with low complexity and high performance are promising for power/energy-sensitive UWOC systems.