Some investigation on optimized localization and routing performance in wireless sensor networks

Abstract

Nowadays, Wireless Sensor Networks (WSNs) have found extensive application in various domains for multi-disciplinary activities, for example tracking of soldiers in the war field, weather monitoring, and patient monitoring etc. WSN can be formed by deploying small size, low power heterogeneous sensor nodes. In sensor networks sensing, processing and handling communication capabilities require more energy to be expended by the sensor nodes. This indeed limits the battery lifetime of the nodes leading to network failure. Therefore, improving the battery life which leads to network lifetime improvement has become an important research topic. Furthermore, various research contributions have proposed different approaches to maximize the battery life and hence the network life in the recent past. To handle the energy constraint issues and improve the node as well as network lifetime, three important suggestions are proposed. First, the traffic conditions are studied and position details are incorporated effectively among the sensor nodes. The Monte-Carlo Localization (MCL) algorithm which results in better localization with the aid of more number of samples is modified by exploiting genetic algorithm and Least Mean Square (LMS) co-efficient. This approach has resulted in optimum set of samples and less number of communication hops and resolves network traffic, communication overhead and localization in-accuracy issues. However, this approach is found to be efficient for static node architecture and loses its efficiency when extended to dynamic architecture. Second, a re-localization strategy that uses pre- defined time interval to avoid localization error observed in the case of modified MonteCarlo Localization is proposed. Specifically, time bounded re-localization approach that exploits Markov Decision Process (MDP) to estimate the optimum re-triggering time of re-localization determined by Camer Rao Lower Bound (CRLB) is outlined. This algorithm identifies optimum re-localization time and global synchronization time to enhance the data gathering capability with the help of Energy Efficient Data Gathering Pattern (EEDGP) algorithm. Here, node tracking and data gathering lead to increase complexity in dynamic scenarios. To overcome this issue, a distributed and prediction based clustering algorithm that defines next positions of the node and to catch the optimal communication path is suggested This proposal henceforth called as Adaptive Dynamic Clustering (ADC) which minimizes the cluster overhead and improves the three tier routing problem newline