Phoenix: A Weight-Based Network Coordinate System Using Matrix Factorization

@article{Chen2011PhoenixAW,
  title={Phoenix: A Weight-Based Network Coordinate System Using Matrix Factorization},
  author={Yang Chen and Xiao Wang and Cong Shi and Eng Keong Lua and Xiaoming Fu and Beixing Deng and Xing Li},
  journal={IEEE Transactions on Network and Service Management},
  year={2011},
  volume={8},
  pages={334-347}
}
  • Yang Chen, Xiao Wang, Xing Li
  • Published 29 November 2011
  • Computer Science
  • IEEE Transactions on Network and Service Management
Network coordinate (NC) systems provide a lightweight and scalable way for predicting the distances, i.e., round-trip latencies among Internet hosts. Most existing NC systems embed hosts into a low dimensional Euclidean space. Unfortunately, the persistent occurrence of Triangle Inequality Violation (TIV) on the Internet largely limits the distance prediction accuracy of those NC systems. Some alternative systems aim at handling the persistent TIV, however, they only achieve comparable… 

Figures and Tables from this paper

Conditional clustered matrix factorization based network coordinate system
Improving prediction accuracy of Matrix Factorization based Network coordinate systems
TLDR
Conditional Clustered Network Coordinate System is proposed, which divides the space into a number of clusters in a balanced, dynamic, and decentralized way and outperforms Phoenix and Pancake significantly in terms of estimation accuracy, expected time to construct the clusters, and the communication overhead.
NCShield: Protecting Decentralized, Matrix Factorization-Based Network Coordinate Systems
TLDR
This paper provides the first study on attacks toward MFNC systems, and proposes a trust and reputation-based approach called NCShield to counter such attacks, which is fully decentralized and can easily be customized.
Self-Stabilized Distributed Network Distance Prediction
TLDR
RMF is proposed, a relative coordinate-based distributed sparse-preserving matrix-factorization method to provide guaranteed stability for the coordinate system, which is self-stabilizing, guaranteeing to converge and not interfered by inaccurate coordinates.
NCShield: Securing decentralized, matrix factorization-based network coordinate systems
TLDR
This paper provides the first study on attacks toward MFNC systems, and proposes a decentralized trust and reputation approach, called NCShield, to counter such attacks, which is able to distinguish between legitimate distance variations and malicious distance alterations.
Network coordinate system using non-negative matrix factorization based on KL divergence
TLDR
The results show that the delay prediction algorithm based on KL-NMF network coordinate system not only reduce the error to reproduce delay matrix, but also improve the network node delay prediction accuracy.
DISCS: A Distributed Coordinate System Based on Robust Nonnegative Matrix Completion
TLDR
This paper presents a novel distributed network coordinate system (DISCS) that utilizes a limited set of distance measurements to achieve high-precision distance prediction at a fast convergence speed and finds that DISCS outperforms the state-of-the-art NCS schemes in terms of prediction precision and convergence speed.
TSMF: Network Latency Estimation using Matrix Factorization and Time Series Forecasting
TLDR
TSMF is devised, a novel NCS method based on an ensemble of Time-Series Forecasting and Matrix Factorization that outperforms Euclidean and MF-based NCS's with up to 6× less relative error in predicting end-to-end latencies.
RNC: A high-precision Network Coordinate System
TLDR
This paper presents a novel distributed network coordinate system based on Robust Principal Component Analysis, RNC, that uses a few local distance measurements to calculate high-precision coordinates without convergence process and indicates that RNC outperforms the state-of-the-art NCS schemes.
WDCS: A Weight-Based Distributed Coordinate System
TLDR
WDCS is proposed, a novel distributed NCS scheme that uses a well-designed algorithm to select an appropriate set of landmarks and employs the W_RUN_PACE method to eliminate the impact of measurement errors or outliers and outperforms the state-of-the-art NCS schemes in terms of prediction accuracy and convergence speed.
...
1
2
3
4
5
...

References

SHOWING 1-10 OF 68 REFERENCES
Network Distance Prediction Based on Decentralized Matrix Factorization
TLDR
Decentralized Matrix Factorization (DMF) achieves competitive accuracy with the double advantage of having no landmarks and of being able to represent distance asymmetries and TIVs.
IDES: An Internet Distance Estimation Service for Large Networks
TLDR
A model for representing and predicting distances in large-scale networks by matrix factorization is presented which can model suboptimal and asymmetric routing policies, an improvement on previous approaches and a scalable system is designed and implemented that predicts large numbers of network distances from limited samples of Internet measurements.
Taming the triangle inequality violations with network coordinate system on real internet
TLDR
This work designs an NC system with a hierarchical architecture, motivated by the natural idea of partitioning the three TIV links into different autonomous NC systems, in order to make as many as TIVs inherently embeddable in metric space.
Measurement Manipulation and Space Selection in Network Coordinates
  • C. Lumezanu, N. Spring
  • Computer Science
    2008 The 28th International Conference on Distributed Computing Systems
  • 2008
TLDR
This paper adapts the Vivaldi algorithm to use Hyperbolic space for embedding and evaluates both Euclidean andHyperbolic Vivaldi on seven sets of real-world latencies, showing that node filtering and latency distributions can significantly influence the accuracy of the predictions.
Vivaldi: a decentralized network coordinate system
TLDR
Vivaldi is a simple, light-weight algorithm that assigns synthetic coordinates to hosts such that the distance between the coordinates of two hosts accurately predicts the communication latency between the hosts.
Predicting Internet network distance with coordinates-based approaches
  • T. Ng, Hui Zhang
  • Computer Science
    Proceedings.Twenty-First Annual Joint Conference of the IEEE Computer and Communications Societies
  • 2002
TLDR
This work proposes using coordinates-based mechanisms in a peer-to-peer architecture to predict Internet network distance (i.e. round-trip propagation and transmission delay), and proposes the GNP approach, based on absolute coordinates computed from modeling the Internet as a geometric space.
Pomelo: accurate and decentralized shortest-path distance estimation in social graphs
TLDR
Pomelo is proposed, which calculates the graph coordinates in a decentralized manner and introduces the novel partial BFS, which greatly outperforms Orion in estimation accuracy while maintaining the same computational overhead.
On the accuracy of embeddings for internet coordinate systems
TLDR
This work defines several new accuracy metrics that attempt to quantify various aspects of user-oriented quality in Internet coordinate systems and indicates that their quality is not as high as that suggested by the use of absolute relative error.
Fundamental Effects of Clustering on the Euclidean Embedding of Internet Hosts
TLDR
This paper investigates why the Euclidean embedding shows reasonable accuracy despite the prevalence of TIVs, focusing in particular on the effect of clustering among Internet hosts, and explains why random selection of a large number of landmarks improves the accuracy.
Handling node churn in decentralised network coordinate system
TLDR
The design and implementation of Myth, a decentralised and fast convergence NC system, is proposed, which introduces the merit of Landmark-based NC system to shorten convergence time in Vivaldi with slight extra overhead and outperforms Vivaldi a lot under node churn.
...
1
2
3
4
5
...