Improving Life Time of Wireless Sensor Networks Using Neural N etwork Based Classification Techniques With Cooperative Routing
Sudhir Gangadharrao Akojwar, Rajendra M Patrikar · 2008
Wireless Sensor Networks are design with energy c onstraint. Every attempt is being made to reduce the energy consumption of the wireless sensor node. Communication amongst nodes consumes the largest part of the energy. The paper focuses on use of classification techniques using neural network to reduce the data traffic from the node and there by reduce energy consumption. The sensor data is classified using ART1 Neural Network Model. Wireless sensor network populates distributed nodes. The co- operative routing protocol is designed for communication in a distributed environment. In a distributed environment, the data routing takes place in multiple hops and all the nodes take part in communication. This protocol has been designed for wireless sensor networks. This ensures uniform dissipation of energy for all the nodes in the whole network. Directed diffusion routing protocol is implemented to carry out performance comparison. The paper discusses classification technique using ART1 neural network models. The classified sensor data is communicated over the network using two different cases of routing: cooperative routing and diffusion routing. Ptolemy-II-Visual Sense is used for modeling and simulation of the sensor network. Lifetime improvement of the WSN is compared with and without classification using cooperative routing and diffusion routing.