IP assignment for efficient NoC-based system design using multi-objective particle swarm optimisation

Abstract

Network-on-chip (NoC) is considered the next generation of communication in embedded system. In this case, an application is implemented by a set of collaborative intellectual propriety blocks (IPs). The selection of the most suited block from a library of IPs is an NP-complete problem. In this paper, we use multi-objective particle swarm optimisation (MOPSO) to yield the best selection of IP to implement efficiently a given application on a NoC infrastructure. In this purpose, MOPSO is exploited to obtain an assignment that minimises the requirements for power, hardware area and the total execution of the application. We show that the achieved solutions are better that those obtained by other multi-objective optimisation algorithms.