Traffic rate network tomography with higher‐order cumulants

Network tomography aims at estimating source-destination traffic rates from link traffic measurements. This inverse problem was formulated by Vardi in 1996 for Poisson traffic over networks operating under deterministic as well as random routing regimes. In this paper we expand Vardi's second-order moment matching rate estimation approach to higher-order cumulant matching with the goal of increasing the column rank of the mapping and consequently improving the rate estimation accuracy. We develop a systematic set of linear cumulant matching equations and express them compactly in terms of the Khatri-Rao product. Both least squares estimation and iterative minimum I-divergence estimation are considered. We develop an upper bound on the mean squared error (MSE) in least squares rate estimation from empirical cumulants. We demonstrate for the NSFnet that supplementing Vardi's approach with third-order empirical cumulant reduces its averaged normalized MSE relative to the theoretical minimum of the second-order moment matching approach by about 12%-18%. This minimum MSE is obtained when Vardi's second-order moment matching approach is based on the theoretical rather than the empirical moments.

[1]  Luiz Henrique Bonani.  Modeling an Optical Network Operating with Hybrid-switching Paradigms , 2016 .

[2]  J. Y. Yen,et al.  Finding the K Shortest Loopless Paths in a Network , 2007 .

[3]  Donald F. Towsley,et al.  Multicast-based inference of network-internal delay distributions , 2002, TNET.

[4]  James J. Iannone,et al.  An EM Approach to OD Matrix Estimation , 1998 .

[5]  Minghua Chen,et al.  Passive Network Tomography for Erroneous Networks: A Network Coding Approach , 2009, IEEE Transactions on Information Theory.

[6]  Paul Barford,et al.  Toward the Practical Use of Network Tomography for Internet Topology Discovery , 2010, 2010 Proceedings IEEE INFOCOM.

[7]  Jian Ni,et al.  Efficient and Dynamic Routing Topology Inference From End-to-End Measurements , 2010, IEEE/ACM Transactions on Networking.

[8]  Robert D. Nowak,et al.  Multiple-Source Internet Tomography , 2006, IEEE Journal on Selected Areas in Communications.

[9]  Michael A. West,et al.  Bayesian Inference on Network Traffic Using Link Count Data , 1998 .

[10]  William H. Richardson,et al.  Bayesian-Based Iterative Method of Image Restoration , 1972 .

[11]  Bin Yu,et al.  Maximum pseudo likelihood estimation in network tomography , 2003, IEEE Trans. Signal Process..

[12]  Donald F. Towsley,et al.  Multicast-based inference of network-internal loss characteristics , 1999, IEEE Trans. Inf. Theory.

[13]  Brian L. Mark,et al.  Delay Network Tomography Using a Partially Observable Bivariate Markov Chain , 2017, IEEE/ACM Transactions on Networking.

[14]  Robert Nowak,et al.  Network Tomography: Recent Developments , 2004 .

[15]  Y. Vardi,et al.  Applications of the EM Algorithm to Linear Inverse Problems with Positivity Constraints , 1996 .

[16]  Joseph A. O'Sullivan,et al.  Deblurring subject to nonnegativity constraints , 1992, IEEE Trans. Signal Process..

[17]  Charles L. Lawson,et al.  Solving least squares problems , 1976, Classics in applied mathematics.

[18]  B. Yu,et al.  Time-varying network tomography: router link data , 2000, 2000 IEEE International Symposium on Information Theory (Cat. No.00CH37060).

[19]  R. Kumar,et al.  Practical Beacon Placement for Link Monitoring Using Network Tomography , 2006, IEEE Journal on Selected Areas in Communications.

[20]  Kin K. Leung,et al.  Efficient Identification of Additive Link Metrics via Network Tomography , 2013, 2013 IEEE 33rd International Conference on Distributed Computing Systems.

[21]  L. Shepp,et al.  Maximum Likelihood Reconstruction for Emission Tomography , 1983, IEEE Transactions on Medical Imaging.

[22]  Avinatan Hassidim,et al.  Topology discovery of sparse random graphs with few participants , 2011, SIGMETRICS '11.

[23]  I. Csiszár.  Why least squares and maximum entropy? An axiomatic approach to inference for linear inverse problems , 1991 .

[24]  Y. Vardi,et al.  Network Tomography: Estimating Source-Destination Traffic Intensities from Link Data , 1996 .

[25]  Martin L. Hazelton,et al.  Network tomography for integer-valued traffic , 2015, 1501.02514.

[26]  G. Michailidis,et al.  Identifiability of flow distributions from link measurements with applications to computer networks , 2007 .

[27]  Yajun Wang,et al.  Network Topology Inference Based on End-to-End Measurements , 2006, IEEE Journal on Selected Areas in Communications.

[28]  Morteza Mardani,et al.  Estimating Traffic and Anomaly Maps via Network Tomography , 2014, IEEE/ACM Transactions on Networking.

[29]  Robert D. Nowak,et al.  Network delay tomography , 2003, IEEE Trans. Signal Process..

[30]  R. Fisher.  The Advanced Theory of Statistics , 1943, Nature.

[31]  Benxiong Huang,et al.  A Survey on Network Tomography With Network Coding , 2014, IEEE Communications Surveys & Tutorials.

[32]  J. Mendel,et al.  Time and lag recursive computation of cumulants from a state-space model , 1990 .