GPSO Hybrid Algorithm for Rectilinear Steiner Tree Optimization

Abstract

Routing is one of the most important stages in the physical layer design of VLSI circuits. The rapid advancement in VLSI technology has necessitated a greater need to optimize the wire length of the circuits i.e. optimize the routing plan. The floor planning and placement strategies manifest a significant effect on the routing plan of a VLSI circuit which in turn requires routing cost optimization. This problem becomes complex when the number of terminal points increases which turns it into an NP-complete problem. In this context, this paper proposes a hybrid metaheuristic optimization algorithm, the Gradient-PSO (GPSO), which hybridizes Gradient Descent and Particle Swarm Optimization for optimization of Rectilinear Steiner Minimal Tree (RSMT).