Binary Particle Swarm Optimization Algorithm for Functional Partitioning of Embedded Systems

Hardware software partitioning deals with the decision to partition a system description to be more suited to be implemented in special purpose hardware or software running on a standard processor. This is the key task of hardware software co-design, as the decision made at the early stage of the design process impact directly on the performance and cost of the system. This paper presents a novel application of Binary Particle Swarm optimization (BPSO) algorithm for hardware software partitioning. The algorithm operates on functional blocks for designs represented as Directed Acyclic Graph (DAG) with the objective to obtain a Hardware or Software implementation that meets performance requirements with a reduced design cost. Test problems are constructed randomly and the optimal solutions obtained from BPSO algorithm are compared with the optimal solutions obtained from traditional genetic algorithm. Experimental results show that BPSO is capable of finding optimal solutions very fast.