一种大规模IP网络多链路拥塞推理算法

发布时间：2018-02-25 03:30

本文关键词： 拥塞链路推理 tomography 贝叶斯网模型拉格朗日松弛贝叶斯最大后验(BMAP)准则　出处：《软件学报》2017年07期 　论文类型：期刊论文

【摘要】：基于最小集覆盖理论的拥塞链路推理算法,仅对共享瓶颈链路进行推理,当拥塞路径存在多条链路拥塞时,算法的推理性能急剧下降.针对该问题,提出一种基于贝叶斯最大后验(Bayesian maximum a-posterior,简称BMAP)改进的拉格朗日松弛次梯度推理算法(Lagrange relaxation sub-gradient algorithm based on BMAP,简称LRSBMAP).针对推理算法中链路覆盖范围对算法推理性能的影响,以及探针部署及额外E2E路径探测发包的开销问题,提出设置度阈值(degree threshold value,简称DTV)参数预选待测IP网络收发包路由器节点,通过引入优选系数?,在保证链路覆盖范围的基础上,兼顾开销问题,确保算法的推理性能.针对大规模IP网络多链路拥塞场景下,链路先验概率求解方程组系数矩阵的稀疏性,提出一种对称逐次超松弛(symmetry successive over-relaxation,简称SSOR)分裂预处理共轭梯度法(preconditioned conjugate gradient method based on SSOR,简称PCG_SSOR)求解链路先验概率近似唯一解的方法,防止算法求解失败.实验验证了所提算法的准确性及鲁棒性.
[Abstract]:A congestion link reasoning algorithm based on minimum set coverage theory only inferences for shared bottleneck links. When congestion paths are congested with multiple links, the reasoning performance of the algorithm drops sharply. An improved Lagrange relaxation sub-gradient algorithm based on BMAP-based Lagrangian relaxation sub-gradient algorithm based on BMAP-based Bayesian maximum a-posteriori (BMAPs) is proposed. As well as the overhead of probe deployment and extra E2E path detection, this paper puts forward setting threshold and degree threshold value (DTV) parameters to pre-select the IP network transceiver router node to be tested, and introduces the optimal selection coefficient. On the basis of ensuring the coverage of the link, taking into account the problem of overhead and the reasoning performance of the algorithm, the priori probability of the link is used to solve the sparse coefficient matrix of the equations for the multi-link congestion scenario in large-scale IP networks. In this paper, a symmetric successive over-relaxation successive over-relaxation (SSOR) preconditioned conjugate gradient method based on SSOR method is proposed to solve the approximate unique solution of the priori probability of the link. The accuracy and robustness of the proposed algorithm are verified by experiments.
【作者单位】：郑州航空工业管理学院电子通信工程学院;西北工业大学电子信息学院;中国人民解放军32147部队;
【基金】：国家重点基础研究发展计划(973)(2012CB315901,2013CB329104) 河南省高等学校重点科研项目(18A510019)~~
【分类号】：TP393.06

【相似文献】