This paper presents an application-specific, heterogeneous multiprocessor synthesis system, named HeMPS, that combines a form of Evolutionary Computation known as Differential Evolution with a scheduling heuristic to search the design space efficiently. We demonstrate the effectiveness of our technique by comparing it to similar existing systems. The proposed strategy is shown to be faster than recent systems on large problems while providing equivalent or improved final solutions.

A formal system-level synthesis model for the concurrent object-oriented design of parallel computer systems, called Multi-token Object-oriented Bi-directional net (MOBnet), is proposed. The MOBnet model extends the standard Petri net by defining (1) multiple tokens to represent different kinds of synthesis control information, (2) object-oriented nodes (places) to denote the system parts under synthesis, and (3) bi-directional arcs to model the design completion check and synthesis rollback operations. In this paper, we first show that MOBnet can serve as a pre-fabrication design methodology analysis tool in ways such as class hierarchy construction, design specification comparison, reachability analysis, and concurrent process management and analysis. We then formally prove MOBnet to be a valid model for concurrent synthesis and give experimental application examples to verify. Finally, solution schemes for the design completion check and synthesis rollback problems are formally validated by analyzing the dynamic behavior of MOBnet, and experimentally illustrated through examples.