Complex Domain Adaptive System Identification Using Sparse Affine Projection Normalized Correlation Algorithms Under Impulsive Noises

Sparse adaptive filters are used extensively for enhancing the filter performance in a sparse system. The affine projection algorithm (APA) is effective in improving the convergence speed for strongly correlated input signals, but it is very sensitive to impulsive noise. Normalized Correlation Algorithm (NCA) is robust in impulsive noise environments. The affine projection normalized correlation algorithm (AP-NCA) used in complex-domain adaptive filters, combines the benefits of APA and NCA and it does not take into account the underlying sparsity information of the system. In this paper, we develop sparse AP-NCA algorithms to exploit system sparsity as well as to mitigate impulsive noise with correlated complex-valued input. Simulation results show that the proposed algorithms exhibit better performance than the AP-NCA for a sparse system. The robustness of these algorithms is evaluated in terms of Mean square error (MSE) performance in the adaptive system identification context.