Lifetime maximization in distributed sensor network with event-triggered adaptive filtering

This paper considers the problem of maximizing the lifetime of a wireless sensor network (WSN) that is deployed to track a parameter vector, subject to a mean-squared-deviation (MSD) constraint. We employ an event-triggered approach using the set-membership normalized least mean-squared (SM-NLMS) adaptive filter in each node wherein the parameter estimate is updated, and hence transmitted to a fusion node (FN), only when the estimation error exceeds a predefined threshold, which is defined to be a triggering event. This problem of maximizing network lifetime (NLT) is posed as an optimization problem which turns out to be non-convex. We propose a suboptimal iterative solution for this problem and show through simulations that the NLT is enhanced significantly compared to a time-triggered approach such as NLMS.