MOBA: Multi Objective Bat Algorithm for Combinatorial Optimization in VLSI

Abstract It has to be conceded that the complexity of any integrated chip gets abated despite of the increased complexity. Notwithstanding, the perplexity of computing escalates exponentially. Thus computational optimization scores in the scene wherein several conflicting objectives has to be optimized without any compromise. Several years, scientists have solved any optimization problem by cogitating and considering it to be a Single Objective Optimization Problem (SOOP). However, mathematics has eulogized Multi Objective Optimization (MOO) methodologies to solve the conflicting tradeoffs. Besides, there is an instigation to converge our research to biologically inspired metaheuristics to solve optimization problems as it is evident from the anthologies and research archive that these heuristics perpetually doing well in solving optimization problems and that too in extension MOO problems. Researchers are captivated in the field to observe the perplexed processes of nature and mimic it solve optimization problems. In this research work, we have contemplated on the Multi Objective Bat Algorithm (MOBA) a biologically inspired metaheuristics and have successfully applied to solve the problem of floorplanning in VLSI design. The peculiar character of echolocation of microbats are being mimicked and applied to solve the problems in VLSI design. The intriguing attributes of Bats; how it strives to take up its prey, are fathomed out and been adopted in problem solving. The problem is considered as a MOO problem wherein equal importance were given to wirelength minimization and dead space minimization. The results are discussed and rivalled with several other bio-inspired algorithms and are portrayed. To comprehend, the MOBA worked well in VLSI floorplanning optimization wherein the problems were considered with a single objective and multi objective fashion.