On Channel Estimation Based on Compressed Sensing for OFDM UWOC Systems

Compressed sensing (CS) has been widely used for channel estimation especially optical wireless communication (OWC) systems. However, the prior works suffer high complexity in pilot insertion for optical orthogonal frequency division multiplexing (OFDM) schemes. In this paper, we propose a novel CS-based channel estimation algorithm using equispaced pilots to reduce the complexity and setting a noise threshold to remove the useless channel taps. We validate the proposed algorithm for underwater wireless optical communication (UWOC) systems. Numerical results suggest that the proposed channel estimation algorithm performs well in terms of estimation accuracy under low pilot allocation complexity.