A novel approach to discrete Particle Swarm Optimization for efficient routing in VLSI design

Meta-heuristic algorithms have a great impact in the wire length minimization at the stage of Global Routing in electronic design automation (EDA). Particle Swarm Optimization (PSO) has already been proved as a good tool in wire length optimization. Investigations carried out regarding the parameter tuning of Discrete PSO (DPSO), develops two variants of DPSO namely Linearly decreasing inertia weight DPSO and self tuned acceleration constant DPSO which exhibit better results than before. Here in this paper, we propose some other variants of DPSO and analyze the results for wire length optimization. The result establishes the fact of even better optimization along with better convergence rate, stability and lesser time complexity. The technique of incorporating mutation process in the proposed DPSO algorithm is highly successful in the Global Routing problem in EDA.