Abstract: This paper presents an optimization mathematic model of load distribution based on integer programming, and proposes an optimization method based on genetic algorithm which is used for solving problems of balanced network flows with high complexity computation multi-constraint. The algorithm carries on the entire spatial parallel search and employs natural number coding to reduce coding space and searching space, and chooses crossover mutation operator with strategies of preserving the optimization individual, it improves its efficiencies and optimization performances by merging automatic information into it which guides the searching direction and efficiencies. The simulation results verify that the algorithm is efficient and evidently improves the status that traditional SPF route algorithm tends to unbalanced network load distribution.

Key words: Traffic engineering, Balanced network load, Genetic algorithm

摘要： 基于整数规划的多限制条件下流量分布优化数学模型，该文设计了一种基于遗传算法的优化算法来求解计算复杂度高的均衡网络流量问题，进行全空间并行搜索，采用自然数编码及选择交叉变异算子，结合最优个体保存策略进行选择，融入自动信息。仿真结果证明算法是有效的，改善了传统路由算法易于导致的网络流量分布不均衡的状况。

关键词: 流量工程, 均衡网络流量, 遗传算法