VLSI partitioning approximation algorithms based on multicommodity flow and other techniques

In this thesis, we present approximation algorithms for partitioning problems on graphs and hypergraphs. The problems find applications in VLSI layout, design for testability and rapid prototyping. In some of the algorithms we employ multicommodity flow techniques. The time required to approximate the multicommodity flow problems determines the time complexities of the partitioning algorithms. We present fast approximations for the multicommodity flow problems. Finally, we present an efficient heuristic for one of the partitioning problems.