基于PBIL进化算法的ATM虚通道路由规划

合理规划ATM(异步传输模式)网络的虚通道路由能有效提高网络业务服务质量和网络资源利用率，针对其NP完全特性，提出了一种基于PBIL(Population-Based Incremental Learning)进化算法的ATM虚通道路由规划算法，有效结合了遗传算法和竞争学习算法的特点，仿真表明，相对于文中提到的其它三种算法，该算法解的质量最高，能有效地收敛到问题的次优解，甚至最优解．

PBIL Evolution Algorithm for VP Routes Programming Problem in ATM Networks

Efficiently programming the VP routes in ATM networks can improve the traffic's quality of service and increase the utilization of network resource. Aiming at it's NP-complete characteristic, this paper proposes a PBIL evolution algorithm for VP routes programming, which combining the trait of genetic algorithm and competitive learning. Simulation results show that among the four algorithms mentioned in this paper, the proposed algorithm gets the best solution, and can converge to the sub-optimization even the optimization solution with a high probability.