An Adaptive Data Aggregation Technique for Dynamic Cluster based Wireless Sensor Networks

Sensor data collection through clustering mechanisms has become a common strategy in Wireless Sensor Networks (WSN). Usually in WSNs the clusters are formed dynamically and repeatedly to get uniform utilization of energy. It has been seen in many large scale dynamic cluster based WSNs that the sizes of clusters, in terms of number of nodes, show high variance. This affects the data aggregation mechanism on the cluster heads such that in a large sized clusters the collected data cannot be placed in small and fixed size packets without incurring significant losses. Therefore, we have developed a novel and an adaptive method of data aggregation that exploits the spatial correlation between the sensor nodes. The main feature of our proposed aggregation method is that in addition to reducing the cost of redundant data transfer in the network, it also optimally utilizes the available space in a packet at each cluster head. The simulation results have shown that in the proposed aggregation method the payload size requirement decreases to almost 25% of the non compressed payload. Also, the distortion percentage in the proposed aggregation method decreases by 16% to 41% as compared to the mean aggregation method.