Evaluating Interpolation-Based Power Management

Abstract

Power management for WSNs can take many forms, from adaptively tuning the power consumption of some of the components of a node to hibernating it completely. In the later case, the competence of the WSN must not be compromised. In general, the competence of a WSN is its ability to perform its function in an accurate and timely fashion. These two, related, Quality of Service (QoS) metrics are primarily affected by the density and latency of data from the environment, respectively. Without adequate density, interesting events may not be adequately observed or missed completely by the application, while stale data could result in event detection occurring too late. In opposition to this is the fact that the energy consumed by the network is related to the number of active nodes in the deployment. Therefore, given that the nodes have finite power resources, a trade-off exists between the longevity and QoS provided by the network and it is crucial that both aspects are considered when evaluating a power management protocol. In this paper, we present an evaluation of a novel node hibernation technique based on interpolated sensor readings according to these four metrics: energy consumption, density, message latency and the accuracy of an application utilising the data from the WSN. A comparison with a standard WSN that does not engage in power management is also presented, in order to show the overhead in the protocols operation.