MPLS over WDM网络静态业务量疏导算法研究

针对基于波分复用的多协议标记交换(multiprotocol label switching over wavelength division multiplexing,MPLS over WDM)网络的静态业务量疏导,即根据给定的网络拓扑和业务量需求矩阵,进行合理路由和流量分配,从而优化网络性能的问题,提出了一种基于拉格朗日松弛的层间分解算法(Lagrangian relaxation based inter-layer decomposition algorithm,LR-ILDA),将整个优化问题分解成MPLS和WDM层的2个子问题,通过2层数据交互迭代的方式得到整个问题的上、下界,从而精确地估算出问题的最优解.通过仿真结果证明,随着业务量增大,上、下界差距越来越小;对大型网络和大量业务请求,本算法可使上、下界在合理时间内聚合到较小范围内.

Study on static traffic grooming algorithms for MPLS over WDM network

Against the static traffic grooming problem of MPLS over WDM network, inter-layer decomposition algorithm (LR-ILDA) based on Lagrangian relaxation was proposed in this paper to optimize the network performance. Firstly, the optimization problem was divided into two problems in MPLS and WDM layer. Secondly, upper and lower bounds of the problem were calculated by iterating the data of two layers alternately. Finally, the solution of the optimization problem was estimated accurately. Simulation results show that the gap of upper and lower bounds become smaller with traffic increasing, and for large scale network with great deal of traffic, upper and lower bounds can be converged in small range quickly by the proposed algorithm.