Multi Attribute Real Time Traffic Inference Algorithm for Botnet Detection in Mobile Ad Hoc Network

The issue of botnet detection has been talked about in a few articles and there are number of techniques has been examined before. The earlier discussed methods handle the botnet detection in stable/unstable networks with little proximity and produces more false results. To overcome the problem of botnet detection in unstable mobile ad hoc network with dynamic addressing schemes, we propose a multi attribute real time traffic inference model algorithm to perform botnet detection. The method sticks to the basics of unstable network conditions and with the restriction of storage scalability. With this limitation, the nodes maintain small set of trace about the earlier transmission and the node details participated in the transmission. Upon receiving the packets from the neighbor node, the method extracts the features from the packet and performs traffic inference performed based on the other routes available to reach the destination. Also, the method identifies the hop details to identify the presence of botnet. Based on both the results the method eliminates the botnet from the network to improve the performance of the network.