Constrained nondominated neighbor immune multiobjective optimization algorithm for multimedia delivery

Abstract

In recent years, artificial immune system (AIS) algorithms is considered to be an effective method to solve the multiobjective optimization problems (MOPs), such as multimedia delivery problem. Though a decent number of solution algorithms have been proposed for MOPs, far less progress has been made for constrained multiobjective optimization problems (CMOPs), which demands a combination of constraints handling technique and search algorithm, e.g. Nondominated Neighbor Immune Algorithm (NNIA). In this paper, we propose a hybrid constraint handling technique of adaptive penalty function and objectivization of constraint violations. In our approach, the dominant population is updated via a method of objectivization of constraint violations and proportional reduction while a modified adaptive penalty function method based on the structure of the search algorithm (NNIA) is utilized to update the active population. We combine the proposed hybrid constraint handling method with NNIA to form the proposed Constrained Nondominated Neighbor Immune Algorithm (C-NNIA) to address the constrained multiobjective optimization problems. To our knowledge, it is the first time NNIA has been applied as the search algorithm for CMOPs. Numerical simulations indicate that the proposed algorithm outperforms the current state-of-the-art algorithms, i.e. NSGA-II-WTY, in both convergence and diversity.