Examination of multi-objective Genetic Algorithm using the concept of a Peer-to-Peer network

The characteristics of a network of Peer-to-Peer Evolutionary Algorithms (P2P EA), which are parallel genetic algorithms, are discussed. We applied the concept of P2P EA, which is a single-objective optimization method, to a multi-objective Genetic Algorithm. To verify the performance of the solution set in the proposed method, we compared the proposed method to the generational multi-objective genetic algorithm NSGA-II. Moreover, to verify the influence of network topology and cache size, we performed the proposed method with three types of network topologies and three cache sizes. Numerical examinations indicated that the difference in network structure has no influence on the solution set, while the cache size affects the solution set.