Probabilistic reasoning in intelligent systems: networks of plausible inference
Probabilistic reasoning in intelligent systems: networks of plausible inference
IDMaps: a global internet host distance estimation service
IEEE/ACM Transactions on Networking (TON)
Network tomography on general topologies
SIGMETRICS '02 Proceedings of the 2002 ACM SIGMETRICS international conference on Measurement and modeling of computer systems
Tomography-based overlay network monitoring
Proceedings of the 3rd ACM SIGCOMM conference on Internet measurement
An algebraic approach to practical and scalable overlay network monitoring
Proceedings of the 2004 conference on Applications, technologies, architectures, and protocols for computer communications
Shrink: a tool for failure diagnosis in IP networks
Proceedings of the 2005 ACM SIGCOMM workshop on Mining network data
Fast maximum margin matrix factorization for collaborative prediction
ICML '05 Proceedings of the 22nd international conference on Machine learning
Collaborative prediction using ensembles of Maximum Margin Matrix Factorizations
ICML '06 Proceedings of the 23rd international conference on Machine learning
ICC'09 Proceedings of the 2009 IEEE international conference on Communications
INFOCOM'10 Proceedings of the 29th conference on Information communications
Efficient active probing for fault diagnosis in large scale and noisy networks
INFOCOM'10 Proceedings of the 29th conference on Information communications
Active collaborative filtering
UAI'03 Proceedings of the Nineteenth conference on Uncertainty in Artificial Intelligence
Adaptive diagnosis in distributed systems
IEEE Transactions on Neural Networks
Packet Loss Rate Prediction Using the Sparse Basis Prediction Model
IEEE Transactions on Neural Networks
Hi-index | 0.00 |
Nowadays the overlay network has greatly improved the performance of the Internet. The overlay network flexibly selects its communication paths and targets and thus can benefit from estimation of end-to-end network performances. For an overlay network with n end hosts, most of the existing systems have to send O (n2) probes into the network and then they calculate the performances of all links. Although these systems to some extent can determine the performances of the links, they have to send plenty of probes into the network, which has generated great traffic and imposed extra overload in the network. In order to address the problem, we propose a new approach based on probe prediction method by which we only need to measure a few probes in the probe set and predict out the responses of the rest probes and then we revise the final prediction results and find out the suspected congested links set. The experiments have shown that we only need to send about 20% of the total probes and infer all the responses of these probes with higher accuracy than ever before.