A hybrid whale optimization algorithm with differential evolution optimization for multi-objective virtual machine scheduling in cloud computing

Virtual machine (VM) scheduling in a dynamic cloud environment is often bound with multiple quality of service parameters; therefore, it is classed as an NP-hard optimization problem. Swarm-based metaheuristics, such as the whale optimization algorithm (WOA), have gained a notable reputation for solving optimization problems. The unique bubble-net hunting behaviour and fast convergence of the algorithm led to the development of a hybrid multi-objective whale optimization algorithm-based differential evolution (M-WODE) technique to solve the VM scheduling problem. The differential evolution (DE) strategy is used to replace the randomly generated solution produced by the WOA to ensure diversity in the solution and to strengthen the local search of the M-WODE. In addition, the DE technique is applied to the Pareto front produced by the WOA to escape local optima entrapment problems. The experimental results showed that the proposed M-WODE outperformed previous algorithms in most cases on makespan and the cost trade-off.