Classification of DoS Attacks in Smart Underwater Wireless Sensor Network

Abstract

Due to deployment in sensitive military areas and other security applications, wireless networks are becoming a famous research spot in the field of computer science. To ratify the security and reliability in such kinds of application intrusion detection system can play an important role. There is a need of intrusion detection system, which has the capability to detect a large number of possible threats in wireless sensor networks. This article contains a customized dataset for smart underwater wireless sensor network that can be categorized into four types of DoS attacks (Blackhole, Grayhole, Flooding and Scheduling attacks). For the experimental purpose, since it is highly used in WSN, Low Energy Aware Cluster Hierarchy protocol has been used in this research work. Vector based forwarding provide scalable routing protocol to deploy smart underwater wireless sensor network. Using NS2 network simulator model a scheme has been defined to collect network traffic and create the dataset. Artificial Neural Network has been applied to train the dataset to classify it into different DoS attacks. Experimental work performed here gives high classification rate and accuracy for mentioning attacks with the help of proposed dataset. In future, suggested method can be useful for more attacks like Sybil/Wormhole presented in datalink layer as DoS attacks.