CIRCUIT PARTITIONING WITH HYBRID PSO ACO APPROACH USING VHDL By Gurpreet K. Walia,Electronics
Abstract
In order to 20 implement a large circuit, it is necessary to break that circuit into smaller 20 circuits because of the various limitations associated with the implementation 20 of large circuits like I/O pin limitations. So, partitioning is necessary as it 20 plays an essential role in the VLSI design process. Partitioning is = basically a 20 design task to break a bigger system into smaller pieces to be implemented on=20 separate interacting components. It also helps to solve difficult and = complex=20 optimization problems. This report deals with the aim to = minimize number=20 of interconnections between the partitions of a VLSI circuit, keeping in = mind=20 all the constraints. The minimized interconnections also reduces the = delay and=20 the interface between the partitions making it easier for independent = design and=20 fabrication. Minimizing the interconnections is also called Mincut = problem. This=20 objective may be optimized using an Optimization technique such as Ant = Colony=20 optimization, Particle Swarm Optimization etc. For this dissertation, = Particle=20 Swarm Optimization(PSO) and Ant Colony Optimization(ACO) algorithms are = used. For testing this method, various = netlists are=20 taken from the Standard ISPD=9298 benchmark. MATLAB was used for generating the = initial=20 partitions, to calculate mincut and for the rest of the work, VHDL was=20 used. We got better results = for=20 certain circuits for min-cut.
