Optimization of security in hybrid networks using computational intelligence

Abstract

Ad-hoc networks are the contributors of the wireless technology that have opened new dimensions in the area of research and technology. Such networks do not rely on preexisting infrastructure for linking files between devices through the internet. The necessary procedures in ad-hoc networks are carried out through the help of all existing available nodes. Owing to this integral shortage of setup, the nodes of ad-hoc networks cannot be deliberated as trustworthy in comparison to devoted infrastructure setup. Therefore, there is a need to secure the structure of ad-hoc networks so that basic operations of systems including routing could be performed efficiently. However, these networks are sensitive to various kinds of assaults. Among the numerous challenges that severely overburdens the resources is the distributed denial of service (DDoS) attack. The repercussion of such attack may further affect routing, Quality of Service (QoS) and security that require optimization. In order to optimize security, network must ensure data confidentiality, data integrity, and availability of services. In this thesis, effect of DDoS attacks on ad-hoc networks has been reviewed and an attempt has been made to provide the optimized solution for enhancing the MANET environment in terms of QoS, performance and security against DDoS attacks. The first phase of the research work aims to provide a detailed comparative analysis among the existing reactive AODV (Adhoc On-demand Distance Vector), SAODV (Secured -AODV) and HWMP (Hybrid Wireless Mesh Protocol) based on QoS parameters with respect to DDoS attacks for WMN (Wireless Mesh Network).The complete analysis has been incorporated using Network Simulator (NS-version 2.3.3) and reveals that HWMP exhibits better performance in terms of various QoS parameters like PDR (Packet Delivery Ratio), PLR (Packet Loss Ratio), jitter, energy consumption, throughput, delay and NRL (Normalized Routing Load). Further, blockchain technology has been incorporated for developing more efficient hybrid and