Enhanced Fuzzy Particle Swarm Optimization Load Distribution (EFPSO-LD) for DDOS Attacks Detection and Prevention in Healthcare Cloud Systems

Abstract

Distributed Denial of Service (DDoS) is an attack that threats the availability of the healthcare related cloud services. In order to assure the each and every one time accessibility of patient’s data, propose a new solution that allows, firstly, the hypervisor to establish credible trust relationships among VMs by considering purpose and personal trust sources and employing vectors to aggregate them. Secondly Enhanced Fuzzy Particle Swarm Optimization (EFPSO) algorithm which guides the hypervisor to determine the optimal loads distribution among VMs in real-time that maximizes DDoS attacks’ detection. EFPSO algorithm which allocates incoming client request to available virtual machines depending on the load i.e. VM with least work load is found and then new request is allocated in the attack detection. The proposed EFPSO algorithm gives the hypervisor with the optimal detection load distribution strategy over VMs that maximizes the detection of DDoS attacks under a limited budget of resources. At finally prevention is performed by using Convex Support Vector Machine (CSVM) classifier. Experimental results are measured in terms of attacks’ detection, false positives, negatives, and CPU, memory during DDoS attacks.