An Evaluation of Effective Intrusion DoS Detection and Prevention System Based on SVM Classifier for WSN

Abstract

Abstract The trending technology of Wireless Sensor Networks (WSNs) are broadly used in this recent earth for a greatest rate of purpose for the better worth of person living however owing to their energy constrains as well as utilization and problems as mentioned above, these WSN has their restriction in their intention as well as their practicality. The wireless sensor networks face main threat challenges such as Denial of Service (DoS) attack, flooding attack, replay attack. In this approach, we proposed an Intrusion Detection (ID) technique using Support Vector Machine (SVM). In the ID algorithm, we employ feature vector to constitute nodes. Such components consist of routing message number, varies destinations in sending and receiving routing packets, and similar source node. The similar source node consists of similar characteristics; the node of the similar category has the same features. Therefore, irregular nodes will be notable. The measurement result shows an introduced scheme attain higher rate of accuracy. Here, Ant colony Optimization (ACO) technique to choose best route between entire adjacent feasible routes from source to receiver node for data communication. The proposed approach enhances the network lifespan as well as discovers intrusion nodes in WSN.