A swarm based global routing optimization scheme

Swarm Intelligence (SI), modelled upon the behaviours of various swarms of animals and insects such as ants, termites, bees, birds, fishes, fireflies, etc. is an emerging area in the field of optimization. SI based algorithms are proclaimed to be robust and efficient optimization tools. This fact is corroborated by a number of practical engineering problems where these algorithms give very satisfactory results. Nowadays VLSI Design has become one of the most intriguing and fervent research field for engineers. Efficient development of a system of a billion chips and blocks on a printed circuit board requires extensive use of optimization in various areas of design such as chip size, separation among components, interconnect length etc. One of the most significant among these is the interconnect wirelength, which determines the overall delay in transmission within the chip. The routing phase in the VLSI Physical Design strives to optimize the interconnect length. Several studies have been and are being conducted to improve the performance of VLSI chips by optimally interconnecting the various components. Various SI based algorithms have already proved their efficiency in this field of routing optimization. In this paper we have proposed a global routing scheme based on contemporary SI algorithms: Firefly Algorithm (FA), and Artificial Bee Colony (ABC) algorithm and have compared the performance of the two. FA produces superior optimization results in comparison to ABC although proving to be quite expensive, computationally.