Develop a Machine Learning Based Algorithm to Detect and Prevent DoS and DDoS Attacks in SDWSN

Abstract

Software Defined Wireless Sensor Network (SDWSN) is a network paradigm recently developed for dynamic and secure control for Internet of Things (IoT) applications. Where it employs the ease of network management and configuration of Software Defined Network (SDN)due to the separation between control and data planes, in order to address the inherent challenges that faced Wireless Sensor Networks (WSN). However, SDWSNs are not immune against the challenges and threats that arise from network intrusions, and there is no active mechanism in place to monitor the network and make it in proactive alert at all times. Also, Denial of Service (DoS) and Distributed Denial of Service (DDoS) attacks are major threats to security in SDWSN networks, as these attacks can overwhelm the network with traffic, preventing legitimate users from accessing the network and disrupt critical network services such as communications, data transmission, control, and monitoring, which cause significant financial and operational loss.

This study proposes implementation of a machine learning-based algorithm to

detect and prevent DoS and DDoS attacks in SDWSNs. The algorithm uses a variety of features, such as packet size, packet rate, packet source, packet destination and a number of statistical calculated features to train a classifier that can distinguish between benign traffic and malicious traffic generated by attacks. The results show that the proposed algorithm can detect and prevent DoS and DDoS attacks with high detection rate, which is reflected in the overall performance of the network. It is also easily scalable and deployable in SDWSNs, which makes a significant contribution to the security of these networks.