A Classification Algorithm to Reduce Data Redundancy in Wireless Sensor Networks

Middleware acts in between the application level and the lower level constructs which solve many wireless sensor network challenges. Still, middleware poses a number of issues such as energy constraint of devices, scalability, mobility, heterogeneity, data aggregation, data redundancy, dynamic networking requirements, security and Quality of Service. If data redundancy is reduced, then data aggregation and energy-related issues in middleware can be solved. An efficient middleware must have the ability to aggregate data from devices in such a way that data redundancy is reduced and hence, the amount of data transmitted to the sink is decreased. Classification algorithms can be used in middleware to reduce the duplication of data. Here, we use k-nearest neighbour algorithm for classification in order to prevent data redundancy in wireless sensor networks.