An iterative greedy algorithm for hardware/software partitioning

Hardware/software partitioning is one of the most important problems in hardware/software codesign. This paper proposes an iterative greedy algorithm for solving hardware/software partitioning. The proposed algorithm iteratively refines a solution by a local search method, a destruction method and a construction method. We show computational results on four benchmarks from the literature. Comparisons with a traditional evolutionary algorithm and an algorithm based on artificial immune principles demonstrate the efficacy of the proposed algorithm in terms of solution quality.