Energy Harvesting in physical sensors using Hybrid Selection of Active nodes in Optimized GAF protocol for Wireless Sensor Networks.

Abstract

Wireless Sensor Network (WSN) has been an active research area and getting popular in wide range of applications. WSN comprises of many sensor nodes that sense, communicate and compute with its battery power. Sensor nodes observe the environmental conditions and convert them into electrical signal. Due to small in size, rapid results and low in price electrochemical sensors are used in a variety of applications. Lithium batteries are alternative to alkaline batteries as it enhances the life and higher current draw. Lithium iron-disulfide batteries are not rechargeable as replacing is most dangerous. Hence, it seems to be important to identify alternate methods to minimize the energy consumption. Geographic Adaptive Fidelity (GAF) is a location aware routing protocol with sleep/active mode. Hence most of the nodes are in sleep mode in GAF protocol, it conserves considerable amount of energy. Selection of active nodes based on the distance which is nearer to the base station may create energy hole problem, as the same set of nodes tend to be selected frequently. On the other hand selecting the active node based on their higher remaining energy may cause unnecessary transmission of data through many nodes. So, we introduced a method of identifying the active nodes with the hybrid factor based on both distance to the base station and higher residual energy. Our proposed protocol namely Hybrid Optimized Geographical Adaptive Fidelity (HOGAF) Protocol, First selects the sequences of nodes to be in active state based on hybrid calibration over residual energy of sensor nodes and distance to Base Station. Next, it implements GAF in which only selected nodes in a grid remains in an active state for data transmission. In GAF, WSN uses location information based GPS, Received Radio Signal Strength or RFID. Our simulation results are compared with DBEA-LEACH and optimized GAF. Results show that our method of selecting the active nodes outperforms the other two methods in terms of network life time and energy consumption