基于遗传算法的互连网络网关与路由方案设计

研究互连网络的网关与路由方案综合优化设计问题，包括：（１）确定将给定的已有网络互连起来所需的网关数及其位置；（２）给网关链路分配容量；（３）进行网内的网际路由选择，目标是要在给定的设置网关预算费用下使得网内的网际平均分组时延最小。该问题的数学模型是一个有着众多约束条件的非线性０—１规划。根据遗传算法的基本思想，设计了一个求解该问题的遗传优化算法，并将其应用于一个实例化设计。

Gateways and route selection design of internet based on GAs

In this paper, we study the joint problem of gateways and route selection design of internet, including: (1) the number of gateways and their locations to interconnect existing networks, (2) the capacity assignments for the gateways, (3) the route selection for the intranet and internet traffic to minimize a linear combination of the average intranet and internet packet delays subject to a cost constraint on the amount to be spent to establish the gateways. This problem is formulated as nonlinear 0- 1 programming. We develop a new optimization method based on genetic algorithms (GAs) to solve this problem. In the computational experiments, the method was shown to be effective in interconnecting two WANs and can quickly obtained the near global optimal solution of the problem.