基于端到端链路利用的网络拓扑发现算法研究

网络拓扑对于网络管理和推测网络内部性能具有重要意义。在分析现有算法局限性的基础上,提出了一种基于端到端测量的网络拓扑发现方法,该方法通过比较链路利用的hamming距离推测网络拓扑,特点是不需要网络内部节点的协助。通过理论比较和分析,证明该方法在推测精度上比现有的方法有显著的提高,最后利用仿真试验验证了该方法的有效性和准确性。

Topology Identification Based on End-to-End Link Utilization

Network topology plays a vital role in many applications such as network measurement and network internal performance inference. By analyzing limitations of the existing algorithms, a new network topology identification algorithm was proposed based on end-to-end measurement. The proposed algorithm identifies network topology using hamming distance of the sequences on delay or not delay of probe packets, which requires no support from internal nodes. Through theoretic analysis and comparison, the conclusion can be reached that the proposed algorithm can significantly improve the inference accuracy. The proposed algorithm was implemented in a simulated network and it validates its accuracy and efficiency.