对现代通信网中具有时延和带宽约束的QoS单播路由问题提出了一种基于改进蚁群算法的路由优化。该算法结合蚁群系统的特点对基本蚁群算法进行了三方面的改进:在信息素初始化中考虑食物源(目的节点)向周围散发气味的过程;将时延和带宽信息加入启发式函数;引入“弱淘汰”法则全局更新信息素。仿真实验表明该算法可以更快地找到满足时延和带宽请求的最小费用路由。
For QoS unicast routing problem under the constraints of delay and bandwidth in modem telecommunication networks, a new routing optimization based on improved Ant Colony Algorithm (ACA) was proposed. Combined with the characteristics of ant colony system, the ACA was improved in three aspects: the process of food giving off smell is considered in the initialization of pheromone; the delay and bandwidth are added to heuristic function; introduces "elimination the worse" to update the global pheromone. Simulation experiments show that the algorithm can find the least cost routing satisfied the request of delay and bandwidth quickly.