Distributed algorithms for maximum lifetime routing in wireless sensor networks

TLDR

A sensor network of energy‑limited wireless nodes is considered. The study proposes distributed algorithms to compute an optimal routing scheme maximizing the time until the first node drains, and extends the formulation to more general lifetime definitions for realistic scenarios. The problem is formulated as a linear program and solved via distributed subgradient algorithms, and the approach extends to many problem variants. The algorithms exhibit low computational complexity, guarantee convergence to an optimal routing scheme maximizing lifetime, and are illustrated by computing an optimal flow in a randomly distributed node network.

Abstract

A sensor network of nodes with wireless transceiver capabilities and limited energy is considered. We propose distributed algorithms to compute an optimal routing scheme that maximizes the time at which the first node in the network drains out of energy. The problem is formulated as a linear programming problem and subgradient algorithms are used to solve it in a distributed manner. The resulting algorithms have low computational complexity and are guaranteed to converge to an optimal routing scheme that maximizes the network lifetime. The algorithms are illustrated by an example in which an optimal flow is computed for a network of randomly distributed nodes. We also show how our approach can be used to obtain distributed algorithms for many different extensions to the problem. Finally, we extend our problem formulation to more general definitions of network lifetime to model realistic scenarios in sensor networks