ABSTRACT
Good clustering can provide critical insight into potential locations where congestion in a network may occur. A natural measure of congestion for a collection of nodes in a graph is its Cheeger ratio, defined as the ratio of the size of its boundary to its volume. Spectral methods provide effective means to estimate the smallest Cheeger ratio via the spectral gap of the graph Laplacian. Here, we compute the spectral gap of the truncated graph Laplacian, with the so-called Dirichlet boundary condition, for the graphs of a dozen communication networks at the IP-layer, which are subgraphs of the much larger global IP-layer network. We show that i) the Dirichlet spectral gap of these networks is substantially larger than the standard spectral gap and is therefore a better indicator of the true expansion properties of the graph, ii) unlike the standard spectral gap, the Dirichlet spectral gaps of progressively larger subgraphs converge to that of the global network, thus allowing properties of the global network to be efficiently obtained from them, and (iii) the (first two) eigenvectors of the Dirichlet graph Laplacian can be used for spectral clustering with arguably better results than standard spectral clustering. We first demonstrate these results analytically for finite regular trees. We then perform spectral clustering on the IP-layer networks using Dirichlet eigenvectors and show that it yields cuts near the network core, thus creating genuine single-component clusters. This is much better than traditional spectral clustering where several disjoint fragments near the network periphery are liable to be misleadingly classified as a single cluster. Since congestion in communication networks is known to peak at the core due to large-scale curvature and geometry, identification of core congestion and its localization are important steps in analysis and improved engineering of networks. Thus, spectral clustering with Dirichlet boundary condition is seen to be more effective at finding bona-fide bottlenecks and congestion than standard spectral clustering.
- R. Andersen, F. Chung, and K. Lang. Local graph partitioning using PageRank vectors. In Proc. IEEE Symposium on Foundations of Computer Science (FOCS'06), pages 475--486, Berkeley, California, Oct. 2006. Google ScholarDigital Library
- A.-L. Barabási and R. Albert. Emergence of scaling in random networks. Science, 286(5439):509--512, Oct. 1999.Google ScholarCross Ref
- F. Chung. Spectral Graph Theory. AMS Press, Providence, RI, 1997.Google Scholar
- F. Chung. Random walks and local cuts in graphs. Linear Algebra Appl., 423(1):22--32, May 2007.Google ScholarCross Ref
- F. Chung, A. Tsiatas, and W. Xu. Dirichlet PageRank and trust-based ranking algorithms. In Proc. Workshop on Algorithms and Models for the Web Graph (WAW'11), pages 103--114, Atlanta, Georgia, May 2011. Google ScholarDigital Library
- P. Erdös and A. Rényi. On the evolution of random graphs. Publ. Math. Inst. Hung. Acad. Sci., Ser. A, 5:17--61, 1960.Google Scholar
- J. Friedman. The spectra of infinite hypertrees. SIAM J. Comput., 20(5):951--961, Oct. 1991. Google ScholarDigital Library
- M. Gromov. Hyperbolic groups. In S. Gersten, editor, Essays in Group Theory, pages 75--263. Springer, New York, 1987.Google ScholarCross Ref
- O. Haggström, J. Jonasson, and R. Lyons. Explicit isoperimetric constants and phase transitions in the random-cluster model. Ann. Probab., 30(1):443--473, Jan. 2002.Google ScholarCross Ref
- E. Jonckheere, P. Lohsoonthorn, and F. Bonahon. Scaled Gromov hyperbolic graphs. J. Graph Theory, 30(2):157--180, Feb. 2008. Google ScholarDigital Library
- E. Jonckheere, M. Lou, F. Bonahon, and Y. Baryshnikov. Euclidean versus hyperbolic congestion in idealized versus experimental networks. Internet Math., 7(1):1--27, 2011.Google ScholarCross Ref
- J. Leskovec, K. Lang, A. Dasgupta, and M. W. Mahoney. Statistical properties of community structure in large social and information networks. In Proc. International Conference on the World Wide Web (WWW'08), pages 695--704, Beijing, China, Apr. 2008. Google ScholarDigital Library
- B. D. McKay. The expected eigenvalue distribution of a large regular graph. Linear Algebra Appl., 40:203--216, Oct. 1981.Google ScholarCross Ref
- O. Narayan and I. Saniee. Large-scale curvature of networks. Phys. Rev. E, 84(6), Dec. 2011.Google ScholarCross Ref
- A. Y. Ng, M. I. Jordan, and Y. Weiss. On spectral clustering: Analysis and an algorithm. In Advances in Neural Information Processing Systems (NIPS'02), pages 849--856, Vancouver, Canada, Dec. 2002.Google Scholar
- S. E. Schaeffer. Graph clustering. Computer Science Review, 1(1):27--64, Aug. 2007. Google ScholarDigital Library
- J. Shi and J. Malik. Normalized cuts and image segmentation. 22(8):888--905, Aug. 2000. Google ScholarDigital Library
- N. Spring, R. Mahajan, and D. Wetherall. Measuring ISP topologies with Rocketfuel. In Proc. ACM SIGCOMM Conference (SIGCOMM'02), pages 133--145, Pittsburgh, Pennsylvania, Aug. 2002. Google ScholarDigital Library
- R. Teixeira, K. Marzullo, S. Savage, and G. M. Voelker. In search of path diversity in ISP networks. In Proc. Internet Measurement Conference (IMC'03), pages 313--318, Miami, Florida, Oct. 2003. Google ScholarDigital Library
- U. von Luxburg. A tutorial on spectral clustering. Statist. Comput., 17(4):395--416, Dec. 2007. Google ScholarDigital Library
- D. Watts and S. Strogatz. Collective dynamics of 'small-world' networks. Nature, 393:440--442, June 1998.Google ScholarCross Ref
Index Terms
- Spectral analysis of communication networks using Dirichlet eigenvalues
Recommendations
Non-local spatial spectral clustering for image segmentation
As one of widely used clustering algorithms, spectral clustering clusters data using the eigenvectors of the Laplacian matrix derived from a dataset and has been successfully applied to image segmentation. However, spectral clustering algorithms are ...
Ensemble learning using three-way density-sensitive spectral clustering
AbstractAs one popular clustering algorithm in the last few years, spectral clustering is advantageous over most existing clustering algorithms. Although spectral clustering can perform well in many instances, the algorithm still has some ...
Non-negative and sparse spectral clustering
Spectral clustering aims to partition a data set into several groups by using the Laplacian of the graph such that data points in the same group are similar while data points in different groups are dissimilar to each other. Spectral clustering is very ...
Comments