POSTER IN IEEE/ACM IPSN 2005 1 Optimizing Timeout-Based Sleep Algorithms

Abstract — For wireless sensor networks that are expected to have a long lifetime while operating on limited energy resources, energy efficiency is a major concern [1]. In particular, it is well known that idle listening of the radio causes significant waste of energy; a power management technique [2] to deal with this problem is to switch to the lower energy consumption SLEEP mode during idle times. The effectiveness of this technique depends on (i) the duration of idle times (ii) the power consumption and performance characteristics of the sensor node and (iii) the sleep schedule/algorithm. A significant body of work focuses today on developing sleep schedules to meet energy and performance requirements. A class of simple, thus easy to implement, algorithms can be based on timeouts: nodes switch from ON to SLEEP, when the radio has been idle for longer than a Timeout period; clearly, the choice of Timeout becomes a critical parameter. In this extended abstract, we propose and analyze fixed timeout algo-rithms for a single sensor node, to minimize energy consumption. First, w