ADMob : An Improved Mechanism for Autonomous Detection of Mobility in Wireless Sensor Networks

Abstract

In many of modern Mobile Wireless Sensor Networks (M-WSN) applications, the use of mobile nodes is considered an important feature for achieving the emergent tasks. Unfortunately, mobility of nodes presents several challenges in the protocol design stage, especially when using highly mobile nodes. Therefore, the design of more efficient mobility handling techniques and algorithms is of great interest to meet the quality-of-service requirements. Indeed, the first step in mobility handling is to allow nodes to be aware of their movements. In this context, because of the lack of efficiency in existing mobility detection solutions, we propose in this paper an improved mechanism, namely ADMob, allowing mobile nodes to detect their mobility autonomously. In fact, by taking advantage of the two well-known metrics, RSSI and LQI, provided by the IEEE802.15.4 compatible wireless radio transceivers, three main contributions are provided in this paper: a low-cost and accurate mobility estimation mechanism based on tracking the distance change between every two peer communicating nodes, an algorithm for optimally detecting the truly mobile nodes in the network. Finally, an adaptive algorithm for dynamically adjusting the parameters of the proposed mechanism according to targeted environments. The proposal is executed at the link layer, then, by using internal trigger messages, the up-layer communication protocols, including MAC, routing and IP protocols, can benefit from the estimated mobility information. Compared with its counterparts, the experiment results revealed that the proposed mechanism has better performance in terms of accuracy, latency, flexibility and adaptability to target environments and especially with no additional energy-hardware costs.