Probabilistic query generation and fuzzy c‐means clustering for energy‐efficient operation in wireless sensor networks

Abstract

SUMMARYDepending upon sensing attributes, wireless sensor networks (WSNs) are classified as event driven, time driven, and query driven. In a given surveillance area, approximation of query generation process using uniform probability mass function (PMF) model seems to be reasonable in aggregate terms based on observations extracted from lifetime span of WSNs. However, owing to random generation aspects of query and the associated temporal variations, the Poisson distribution-based model appears to be more appropriate to resemble the realistic query generation pattern. Invariably, in all the sensor network architectures, the energy management requires an important consideration owing to limited energy resources. For the optimal utilization of energy resources, we propose fuzzy c-means (FCM) algorithm to form clusters in a hierarchical network configuration. Network performance is measured in terms of key performance measures, namely, average residual energy status, critical residual energy status (CRES), and number of network nodes that attain the CRES mark. These performance measures are estimated and analyzed for three different PMF models of query generation namely Uniform, Gaussian and Poisson. The merit of deploying FCM algorithm in terms of maintaining much better energy profile of the entire network is discussed.

Clustered chain founded on ant colony optimization energy efficient routing scheme for under-water wireless sensor networks

“…To achieve energy efficiency, the developers of [20] employ stochastic hill climbing, spatial routing, and the mobile sink technique, in which a sink creates a routing with stochastic hill climbing based on data obtained from cluster heads and sink movement. Fuzzy c-means (FCM) algorithm is used by Pramod and Chaturvedi [21] to build networks, and the network's performance is determined by the energy consumption (residual, critical residual status). Uniform, Poisson, and Gaussian models are used to calculate the energy.…”