An adaptive Tabu Search approach for circuit partitioning

Experimentation indicates that selecting the appropriate parameters that control strategies such as Tabu Search and Genetic Algorithms has a drastic effect on the quality of the final solution. It is important to identify when and how parameters are stable, and devise methods to adjust these parameters depending on problem size and application. This paper introduces a new technique to tune the parameters that control the performance of the Tabu Search algorithm in a more systematic fashion. Results using the adaptive search heuristic indicate that a 10-15% improvement in cut-size are obtained over a traditional Tabu Search method.