Abstract—Mapping is a key issue in NoC system-level design process. The power consumption and delay of a NoC system are primary determined by the mapping result. Since NoC mapping is a NP-Complete problem, which means it beyond the computation ability to obtain the global optimal solution if the size of problem is too large. Approaches in this area focuses on heuristic algorithms. In this paper, a new swarm-intelligence algorithm - Enhanced Chaos Discrete Artificial Bee Colony (ECDABC) algorithm is introduced to solve this problem, which carries out two-step optimization processes. Moreover, the enhanced chaos mechanism in can effectively avoid the defection of premature convergence which are common in most of heuristics. By comparing with two prevalent mapping algorithms – Genetic Algorithm and Particle Swarm Optimazation algorithm, ECDABC algorithm has better performance on the optimized results in the same iterations times, and enhanced chaos mechanism has significant effect.
Index Terms—Networks-on-chip, mapping, multi-objective optimization, artificial bee colony algorithm.
The authors are with the National Key Laboratory of Science and Technology on Communication, UESTC Chengdu, China (Tel: +86- 02861830326, Fax: +86-02861830326, e-mail: firstname.lastname@example.org, email@example.com, firstname.lastname@example.org, email@example.com).
Cite: Lili Pan, Zhe Li, and Xiang Ling, "NoC Multi-Objects Mapping Based On Enhanced Chaos Discrete Artificial Bee Colony Algorithm," International Journal of Future Computer and Communication vol. 1, no. 2, pp. 116-120, 2012.