A conjugate-gradient-based block adaptive algorithm, its applications and performance evaluation

Recently, an adaptive algorithm with desirable properties and useful applications, namely the Optimum Block Adaptive Shifting (OBAS) algorithm has been proposed. This paper proposes a technique to enhance the performance of the OBAS algorithm by using search directions other than the negative gradient directions used in the OBAS algorithm. The proposed directions are based on the conjugate gradient optimization method which leads to a Block Conjugate (BC) algorithm for adaptive filtering. Simulations are performed to compare the BC and the OBAS algorithms. Results of the simulations verify that the BC algorithm converges faster than the OBAS algorithm when the size of the block is relatively large (twice the adaptive filter order) and the signal to noise ratio is relatively high.