VLSI module placement based on probability search space smoothing

Search space smoothing (SSS) was originally proposed to deal only with traveling salesman problem (TSP). In this paper, two smoothing effects, local smoothing effect and global smoothing effect are distinguished firstly by analysis and experiments. By making use of local smoothing effect in search process, a local search that can make use of local smoothing effect is designed for search space smoothing method. A search space smoothing method that can take advantage of local smoothing effect is named as probability search space smoothing. A VLSI module placement algorithm based on probability search space smoothing is proposed and is test by using MCNC benchmarks ami33 and ami49, experimental results are better than the published results so far.