基于改进蚁群算法的网络负载均衡路由优化

最短路径优先的路由控制策略在解决突发业务模式下的拥塞问题上存在不足，针对该问题，采用随机选择尺度因子实现负载均衡，增加路由器向所有相邻路由器转发分组的机会。设计路由算法模拟器，完成改进蚁群网络路由优化算法在4种典型网络上的仿真。以CHINANET为例的对比实验表明，在重负载情况下，改进的蚁群网络路由优化算法能获得较高的吞吐量、较低的平均时延与丢包率。

Routing Optimization for Network Load Balance Based on Improved Ant Colony Algorithm

The shortest path first algorithm always makes the routers select the shortest path to forward the packets, thus often leads to network congestion. Aiming at this problem, this paper uses a random factor to scale the probability of routing select for the purpose of realizing load balance and increasing the probability of forward the packets to all adjacent routers. It designs a routing algorithm simulator in order to validate the improved ant colony routing algorithm, and four typical network scenes are tested in the simulator. CHINANET is taken as an example, for the heavy load condition, the simulative result presents that the improved algorithm shows higher throughput, lower average latency and packet loss ratio.