A time improvement PSO base algorithm using multithread programming

This paper proposes a method to improve the processing time of algorithm that bases on Particle Swarm Optimization (PSO) such as Original PSO, AS-PSO (Ant Supervised by Particle Swarm optimization), ACO (ant colony optimization) by using multithread programming technique. Typically, the PSO Base algorithm creates several particles to find the optimized solution. Each particle owns its local process and resource, but original implementation has been processed in a single thread or sequential computation system. That causes the system to take a longer time to compute when increasing the number of particles. It makes the system slow down when implementing a massive amount of particles and ambiguous to estimated efficiency for each algorithm. This method takes advantage of a multithread computing system for utilizing resources on the system to reduce computation time. The proposed method use thread as particle to manage local resource and implement multithreading system and use thread synchronization system to manage global resource and exchange data between particle. To measure the efficiency of this method Traveling Salesman Problem (TSP) used to be a problem to optimize and compare computation time between the original method and the proposed method. The experimental result show correctness and time improvement of the proposed method compared with the original method.