Optimization Approaches for Core Mapping on Networks on Chip

Abstract

Network on Chip (NoC) has been suggested as an appropriate solution for the communication demands in Systems on Chip (SoCs). Due to the limitation of the resources in NoC, suitable load distribution over limited resources which is known as mapping problem is a challenging issue. This paper proposes a linearized model of quadratic assignment problem (QAP) to map the tasks to the cores not only to minimize the power consumption but also to improve the performance in NoCs. In addition to the proposed linearized model, two meta-heuristic algorithms such as genetic algorithm and simulated annealing are suggested to solve the considered mapping problem as QAP. The extensive computational study over the well-known benchmarks of the literature of NoCs proves that the proposed linearized formulation performs better than the meta-heuristic approaches in terms of power dissipation and performance of the NoCs.