Optimized load balance scheduling algorithm

Abstract

The cloud computing environment faces several challenges as a federation of clouds, controlling the traffic flow, scalability, and balancing the load on virtual machines that are considered the most crucial issue due to their impact on the execution time, resource utilization, and cost. This paper is interested in some of the existing algorithms that distribute the workload evenly. These algorithms aim to avoid the blind assignment that often results in some over-loaded servers while another node might be under-loaded. In this work a combination of two inspired metaheuristic algorithms BAT and cuckoo search was proposed; the first algorithm can utilize fast exploration using global search, the latter algorithm can avoid trapping into BAT local optimum problem using levy flight with a far random walk. Additonaly, the proposed algorithm could be used to mitigate distributed denial of service (DDoS) attack that aims to cause endless load on the servers and stop the service. Experimental results for five virtual machine (VM), ten VM, with the varying number of tasks showed that the proposed algorithm has better resource utilization and less makespan time in almost all the cases.