AN EFFICIENT SPATIAL AND TEMPORAL CO-RELATION MODEL FOR WIRELESS SENSOR NETWORKS

Abstract

A wireless sensor network (WSN) is set of distributed autonomous devices called sensors for monitoring properties like motion, vibration, pressure, temperature, sound, pollutants at different locations. WSN is a set of sensors, located in a remote location equipped with sensing and computational and capabilities. Along with sensors, each node in WSN is mounted with a radio transceiver, an energy source (battery) and a small microcontroller. These low power and low cost devices form a set for monitoring the area of interest. Using the collaboration of sensors, the Wireless Sensor Nodes collects and sends information about the area of interest (e.g. surveillance, rain, humidity, temperature etc.) to the Processing Node. Then received information is processed at Processing Node and used by users. Due to limited energy capacity of nodes, network overall lifetime will not be long. Giving continuous power supply to nodes or changing the used battery is a difficult task in such remote location. This could interrupt or badly affect the flow of information from area of observation. Major portion of the energy consumption of sensor node in the sensor network is happens in the information flow (transmission and reception), and only very few portion in other activities (clustering, broadcasting etc). So to increase the network lifetime once can target to make the sensor nodes into sleep mode as frequent as possible without significant loss in information. If sensor nodes are deployed densely in the area of observation there are chances of getting similar information from the neighboring nodes. Also if the frequency of sending information to the processing node by the DELHI TECHNOLOGICAL UNIVERSITY Page v sensor node is high, again there are chances of observing similar information by the same node. This means observed information can be Spatially and Temporally correlated information will be observed by the deployed nodes. Thus Spatial and Temporal correlation provides a basis for proposing and developing techniques for energy saving of sensor nodes. We are proposing a hybrid approach to achieve target of battery saving and the approach is based on Spatial and Temporal Co-relation among various sensor nodes. On sensing a physical phenomenon in an area all nodes start sending observed information to the Processing Node. In a highly dense deployed sensor networks there are lot of chances that these nodes will be sending Spatially and Temporal correlated information. Thus it is not necessary that all deployed nodes send information to the Processing Node, and some nodes which are sending spatially correlated information can be made to sleep mode. Network is divided into number of clusters, which we have taken as 5% of the total number of nodes of a network. Nodes are assigned to the cluster having minimum distance to the cluster head having maximum energy. The distance between nodes is Euclidean Distance. Simulations results show that our proposed technique which is based on clustering and exploiting Spatial and Temporal Correlation can lower the energy consumption of data transmission and thus increasing node and overall network lifetime.