STATISTICAL PERFORMANCE ANALYSIS OF A FAST STOCHASTIC GRADIENT FOR CONSTRAINED ADAPTIVE TIME-DELAY ESTIMATION

An efficient bi-state stochastic gradient is proposed for spontaneous constrained time delay estimation. The quantized stochastic gradient i s an approximation of the polarity of the instantaneous delay estimati on error. It is adjusted in such a way that it has a much higher proba bility to move in the correct direction at each iteration so as to ena ble a speed-up in the delay estimate to converge to global minimum in steady state. The performance of the delay estimator is evaluated stat istically and an analytical solution for its convergence behavior is e stablished. It is demonstrated that the proposed algorithm has at leas t a two-fold improvement in convergence speed when compared with the c onventional approach, and this is verified by extensive simulation res ults.