ANALYSIS AND EVALUATION OF RANDOM PLACEMENT STRATEGIES IN WIRELESS SENSOR NETWORKS

Abstract

Wireless sensor networks (WSNs) are an attractive choice for many applications where sensors are densely and efficiently deployed in an area of interest. For example, in surveillance applications, where deterministic placement is not applicable, the sensors have to be randomly deployed entirely or partially in an area. Whether random or deterministic, the strategy of deploying the sensors has a large impact on the performance of WSNs. In this paper, we propose a systematic methodology for sensors placement based on several random distributions. The quality of deployment is evaluated using a set of proposed measures, the significance of which highly depends on the WSN application. The impact of these deployment strategies on the network performance is thoroughly studied. Moreover, we propose a new hybrid deployment strategy that aims to achieve the best performance based on the suggested deployment quality measures. Both the placement strategies with the deployment quality measures are evaluated using extensive simulations. The results show that the proposed hybrid strategy outperformed other strategies including uniform, random, Gaussian, and exponential distributions. In addition, the paper identifies the family of applications where each of the quality measures is highly critical.