摘要
针对大规模网络中的流量工程问题,设计了结合遗传算法与模拟退火以及结合遗传算法与爬山法的两种混合算法.仿真实验表明,结合遗传算法和模拟退火的混合算法在整体性能上较单独使用遗传算法有明显的提高,结合遗传算法和爬山法的混合算法在处理超大规模的网络和海量业务要求的流量工程问题时具有时间上的明显优势.
A hybrid algorithm combined by a genetic algorithm and simulated annealing was proposed for large scale traffic engineering problems, which performed much better than the pure genetic algorithm and another hybrid algorithm combined by the genetic algorithm and hill climbing method was devised, which required much less CPU time and was ascendent while the network was very large. The simulation results show the effectiveness of the methods.
基金
国家自然科学基金(70171053
70471033)
中国科学技术大学研究生创新基金(KD200408)资助
关键词
流量工程
混合算法
遗传算法
模拟退火
爬山法
traffic engineering
hybrid algorithm
genetic algorithm
simulated annealing
hill climbingmethod
