Efficiency of Classification Techniques in Detecting Congestion in Wireless Sensor Networks

Abstract: Wireless Sensor Network (WSN) is network of hundreds/thousands of sensor nodes. Each node is capable to sense, process and transmit the environmental information. Congestion occurs in wireless sensor networks when an event occurs. Congestion leads to performance degradation of a system. The data mining techniques help to detect congestion and then it can be mitigated by adjusting transmission rate. In this paper we analyze the efficiency of data mining classification to detect the congestion in WSN. In this paper we have implemented PART, RIPPER and j48 classification algorithms to detect the congestion over the network. For the given data set, it is found that PART algorithm is more accurate than RIPPER and j48 in detecting the congestion.

Keywords: WSN, Data mining, Congestion control, classification, PART, RIPPER, j48 algorithm.