Multi-objective Optimization Research and Applied in Cloud Computing

In many real-life applications, a decision maker often needs to handle different conflicting objectives. Problems with more than one conflicting objective are called multi-objective optimization problems (MOPs). Multi-objective evolutionary algorithms (MOEAs) have been developed for solving MOPs. MOEAs have been shown to perform well on some MOPs with two or three objectives; however, MOEAs have substantial difficulties for tackling MOPs with more than three objectives, often referred to as many-objective problems (MaOPs) nowadays. In my thesis, first, I plan to propose an efficient multi-objective artificial bee colony algorithm based on decomposition for solving MOPs. Then, another effective adaptive many-objective evolutionary algorithm is designed to deal with MaOPs. What's more, based on defining a multi-objective optimization model of task scheduling in cloud computing, I use an improved particle swarm optimization algorithm to solve the model. Finally, I try to establish a many-objective optimization model of offloading in mobile edge computing, and find a suitable many-objective evolutionary algorithm for solving it. The proposed algorithms are compared to several state-of-the-art algorithms on these models. The experimental results will show the efficiency and effectiveness of the proposed algorithms.