Energy efficient reliable transmission using SVM based data aggregation & FEC based loss recovery in heterogeneous wireless sensor networks

In wireless sensor networks, greater energy is spent for communication when compared to computation and storage cost. Overlapping of sensing regions is higher, when the sensors are densely deployed and nodes in the same sensing region transmit the similar information to the base station. Hence the energy is wasted in communication part because of the high redundancy. So we need methods to liberate redundant communication, but event report is the essential aspect in sensor network applications. Thus reliable data transfer is also equally important. In this proposed method we are avoiding redundancy as well as ensuring reliable delivery. The aggregator node makes use of Support Vector Mechanism (SVM) to classify the received data based on the correlation that exists in the data from the various finish points. And it will eliminate the redundant information. Forward erasure coding (FEC) technique is applied for loss recovery. From the results, it is observed that our approach prolong the network lifetime by effectively controls the redundancy and achieves reliable transmission.