Bounding Distributed Energy Balancing Schemes for WSNs via Modular Subgames

Abstract

This paper studies means of archiving energy balance in Wireless Sensor Networks (WSNs) as criterion to increase network lifetime. While protocols for balancing energy through centralized computation have been established, effective solutions for more realistic distributed settings are still an open research field. In this work, we present a novel approach for balancing energy in WSNs in a distributed manner by solving a sequence of modular strategic subgames. In addition, we prove theoretical bounds for its solution quality. To the best of our knowledge, we are the first to explicitly provide theoretical bounds for distributed WSN energy balancing schemes. We do so by formulating two specific routing games in which sensor nodes act as strategic agents with interests in energy balance. The first formulated game is a monotone utility game and by exploiting and adapting some existing results, we can prove a meaningful relative bound for its solution value via our distributed modular subgames scheme under realistic conditions. For the second game we provide an absolute bound. We round up our theoretical results by an experimental evaluation, where our modular subgames scheme shows increased performance compared to state-of-the-art algorithms.