A bias-compensated error-modified normalised least-mean-square algorithm is proposed. The proposed algorithm employs nonlinearity to improve robustness against impulsive measurement noise, and introduces an unbiasedness criterion to eliminate the bias due to noisy inputs in an impulsive measurement noise environment. To eliminate the bias properly, a new estimation method for the input noise variance is also derived. Simulations in a system identification context show that the proposed algorithm outperforms the other algorithms because of the improved adaptability to impulsive measurement noise and input noise in the system.
An efficient variable step-size diffusion normalised least-mean-square algorithm is proposed via a mean-square deviation (MSD) analysis for the distributed estimation. The proposed algorithm has two distinguishing features for computational efficiency. In the adaptation step, an intermittent adaptation rule that dynamically adjusts an update interval is proposed to reduce the redundant updates. In the diffusion step, instead of the existing combination rules, a selection rule is proposed to select the intermediate estimate of the most reliable node among its neighbour nodes for the estimate at each node. Moreover, to achieve both fast convergence rate and low steady-state error, a variable step size is obtained by minimising the MSD.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.