Parallel Applications Mapping onto Network on Chip Based on Heterogeneous MPSoCs Using Hybrid Algorithms

The Contribution

Since the mapping problem is considered as the one of the most design challenges in NOC based heterogeneous MPSoC which affects greatly the final system performance, the design of robust optimizers is required. Several optimization methods such as exact methods and metaheuristics have been proposed in the literature to tackle this problem. Due to today’s high systems complexity, the exact methods have quickly reached their limitations because of their unacceptable runtime. Metaheuristics stand out as highly appropriate approaches to tackle this problem. As stated above, these metaheuristics are classified into S-metaheuristics and P-metaheuristics. These latter, especially Multi-objective Evolutionary Algorithms (MOEA), are widely used and have proved their success in solving this problem. The solution is to further improve these MOEAs using the local search-based metaheuristic called AMOSA and gives new approaches which draw advantages from both metaheuristic types and consequently which supply designers with the best trade-off mapping solutions.