层析成像技术中的自适应网络拓扑推断算法

基于层析成像技术的网络拓扑推断由于采用单一的网络性能参数而在推断时受到网络负载的影响.为了减少单一参数在推断时的限制,提出了一种单播网络中自适应结合多种网络性能参数的拓扑推断算法,该算法能适应不同网络负载的环境并且在应用时仅需适量的计算而无需增加测量流量.分析了自适应拓扑推断算法的正确性并通过NS2进行了仿真,仿真结果表明自适应网络拓扑推断算法能在不同网络负载的环境下准确的推断网络拓扑.

Self adaptive network topology inference algorithm in tomography

There are a few network topology inference techniques based on network tomography, but all of them use only one network performance characteristics, which leads to many limits when the network load is different. In order to reduce the limits of the inference based on one network performance characteristics, a self adaptive network topology inference method is proposed which joines multiple network performance parameters self-adaptively in inference. In applying the self adaptive network topology inference, no additional traffic is needed except for some calculation, and the inference method could be applied in complex networks with different loads. The self adaptive method is analyzed theoretically and validated through simulations by NS2, and the results of simulation illustrate that the self-adaptive inference method could infer network topology correctly when faced with networks whose load changes greatly.