Channel Estimation using SC-PRLS for underwater Communication

Sparseness controlled proportionate recursive least square algorithm (SC-PRLS) was proposed to work with a time-varying sparse environment. PRLS and SC-IPNLMS inspired it. The proportionate matrix distributes gain among filter coefficients during the updating process. The gain distribution coefficient is calculated using a degree of sparseness to deal with varying sparsity. The underwater acoustic (UWA) channel model considered different characteristics of the channel. By changing the channel impairments, UWA channel sparsity can be modified. The modeled UWA Channel was estimated using PRLS and SC-PRLS. The simulation was done in MATLAB for UWA channel estimation. The simulated results validate better tracking accuracy in a variable sparse environment. The convergence rate and steady-state error are both better for SC-PRLS than PRLS. For SC-PRLS, convergence begins at 105 iterations, but for PRLS, it begins at 160 iterations. For PRLS and SC-PRLS, the steady state error is found to be approximately 24.5 dB and 28 dB, respectively. It can be suggested that SC-PRLS performs better in all sparse conditions like sparse, relatively less sparse, and highly sparse environments. SC-PRLS also gives an overall robust performance.