Integrating timed automata into tabu algorithm for HW-SW partitioning

Hardware/software (HW-SW) partitioning is a key problem in codesign of embedded systems, studied extensively in the past. This paper describes a new approach to hardware/software partitioning for synchronous model. We formalize the partitioning problem using timed automata, which captures the key elements of the partitioning. Then the tabu algorithm is applied to timed automata model to search for the solution efficiently. An industrial experiment is conducted to show our approach can handle large applications with hundreds of nodes in task graph effectively