Energy implications of clustering in heterogeneous wireless sensor networks - An analytical view

Abstract

Research on wireless sensor networks has often assumed that nodes are homogeneous. In reality, even homogeneous sensors have different capabilities like different levels of initial energy, drain/depletion rate, etc. This leads to the research on heterogeneous networks where at two or more types of nodes are considered with the network divided into clusters where the more powerful sensors act as cluster-heads and handle resource-demanding tasks. By assuming that the more powerful nodes have much more energy as compared to the other normal nodes, it can lead to the network becoming partitioned when the other nodes deplete their energy; the entire network essentially becomes useless thereafter. A key design implication is that the energy supply built into the powerful nodes need not exceed a threshold, as beyond which, the extra provisioning is redundant. In this paper, we consider an event-detection heterogeneous network with two types of nodes, primarily with different maximum (startup) power levels. A multihop clustering topology is assumed and we analyze the lifetime of the network with respect to the maximum size of the cluster, as well as, the ratio of power levels between the two types of nodes. © 2006 IEEE.