Denial of Service Attack Mitigation Technique for Wireless Networks using Machine Learning Methodology

Abstract

A Wireless Sensor Networks (WSNs) consists of small sensor nodes that collect important information about an area of interest. The ability of these networks to monitor remote and hostile locations has attracted a significant amount of research over the last decades. As a result of this research, WSNs have make their importance in different applications such as transportation, industrial automation, healthcare, habitat monitoring and military surveillance. These networks can be operated in human-unapproachable lands and collect data on an exceptional scale. However, these networks face several technical challenges at the time of deployment as well as operation. Resource limitations such as battery power, inadequate resources for computation arise problems during their operations. One of the major concern which demands proper attention is energy conservation. To extend the lifetime of these networks an energy–efficient routing protocol is very much required. Moreover, the deployment of these networks in a threatening environment and their operation in the presence of error-prone communication links render these networks to several security loopholes. As a result, there is a great need to design a routing protocol which is more energy efficient and robust against energy drainage attacks. Low Energy Adaptive Clustering Hierarchy (LEACH) is a hierarchical-based routing protocol which is considered as an energy-efficient routing protocol. This thesis aims to provide an analysis of LEACH protocol under an energy-drainage attacks and then provide effective solution to detect and then minimizing the effects of these attacks. Selection of LEACH protocol is based on usage of this protocol on industrial level and in terms of energy efficiency LEACH protocol is highly adoptable. Due to these reasons LEACH protocol is also prone to different type of attacks. Although it is an energy-efficient protocol but there are some energy drainage attacks that make this protocol less efficient. We also discuss some energy drainage attacks like Grey-Hole, Scheduling and Vampire Attack on LEACH protocol and proposed a scheme to mitigate these attacks. The energy drainage of the nodes is mitigated by using Cellular Automata (CA) scheme for cluster head selection. The main objective of this thesis work is selection of an optimal cluster heads regarding their trust values, reduces the energy consumption and mitigate from energy drainage attacks associated with malicious CH selection. Only suitable nodes are chosen as CHs to improve usage of the resources depending upon their trust values. We also propose some key ideas about the creation of clusters in network and density of the associated nodes with CH in a cluster according to their positions.