Hybrid Localization in Wireless Sensor Network

Abstract

In wireless sensor network, sensor nodes are deployed in a region where sensor device senses different type of data and send data to sink node for a particular application. In certain application like environment monitoring, forest fire, Industrial monitoring, Trafic Control monitoring, etc measurement of data is meaningless without the knowledge of precise location from where it is obtained. Location based service is highly demanded due to increase of smart devices. Localization is a technique in which unknown node determine it position using different approach. Localization method so far uses basic parameter like signal strength, communication range, angular separation, time difference of arrival, etc. This parameter are highly affected by environmental factors like temperature, humidity, pressure, surrounding noise which leads to error and affect accuracy. Beside centralized approach is suitable for small and sparse network. Distributed approach is suitable for scalable and in large network. This master thesis, we proposed hybrid approach for randomly deployed scalable network that will perform dynamic decision which takes benefit of centralized method as well as distributed method, based on node deployment, energy constrain of node, number of anchor node, signal strength, node connectivity. To reduce environmental factor Filtering process is performed. This will ensure reliable, self localizing, scalable, energy efficient in a heterogeneous network.