A Multi-View Subspace Learning Approach to Internet Traffic Matrix Estimation

Kumar, Awnish; Vidyapu, Sandeep; Saradhi, V. Vijaya; Tamarapalli, Venkatesh

doi:10.1109/tnsm.2020.2983329

Cited by 8 publications

(3 citation statements)

References 31 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…It would be interesting to adapt our NMF based traffic matrix estimation model into a deep NMF model. Other directions of research include the use of other NMF models to perform the traffic flow estimation, such as [48,49], or to improve prediction using data coming from other sources than the traffic flows and then use multi-view techniques such as [21,50].…”

Section: Discussionmentioning

confidence: 99%

“…In [20], Lu et al used multi-fractal discrete wavelet transform (MDWT) to split traffic matrix into different frequency component then train the neural network to predict low and high frequency component of traffic matrix. Kumar et al in [21] proposed a multi-view subspace learning technique for traffic flow estimation. They proposed a novel robust approach to obtain traffic flows from multiple traffic views yielded from rather inexpensive existing methods.…”

Section: Related Workmentioning

confidence: 99%

See 1 more Smart Citation

Structured Nonnegative Matrix Factorization for Traffic Flow Estimation of Large Cloud Networks

Atif,

Gillis,

Qazi

et al. 2021

Preprint

View full text Add to dashboard Cite

Network traffic matrix estimation is an ill-posed linear inverse problem: it requires to estimate the unobservable origin destination traffic flows, X, given the observable link traffic flows, Y, and a binary routing matrix, A, which are such that Y = AX. This is a challenging but vital problem as accurate estimation of OD flows is required for several network management tasks. In this paper, we propose a novel model for the network traffic matrix estimation problem which maps high-dimension OD flows to low-dimension latent flows with the following three constraints: (1) nonnegativity constraint on the estimated OD flows, (2) autoregression constraint that enables the proposed model to effectively capture temporal patterns of the OD flows, and (3) orthogonality constraint that ensures the mapping between low-dimensional latent flows and the corresponding link flows to be distance preserving. The parameters of the proposed model are estimated with a training algorithm based on Nesterov accelerated gradient and generally shows fast convergence. We validate the proposed traffic flow estimation model on two real backbone IP network datasets, namely Internet2 and G ÉANT. Empirical results show that the proposed model outperforms the state-of-the-art models not only in terms of tracking the individual OD flows but also in terms of standard performance metrics. The proposed model is also found to be highly scalable compared to the existing state-of-the-art approaches.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

Structured Nonnegative Matrix Factorization for Traffic Flow Estimation of Large Cloud Networks

Atif,

Gillis,

Qazi

et al. 2021

Preprint

View full text Add to dashboard Cite

show abstract

“…Polverini et al defined an effective criterion based on a flow spread concept that allows to select the optimum set of OD flows to be measured in SDN [16]. Other performance improvements for TM estimation were introduced in references [17]- [19].…”

Section: Related Workmentioning

confidence: 99%

Network Tomography and Partial Least Squares for Traffic Matrix Estimation

Cuberos

Herrera

Wasielewska

et al. 2021

2021 17th International Conference on Network and Service Management (CNSM)

View full text Add to dashboard Cite

The traffic matrix is a useful data structure in network management, monitoring, optimization and traffic forecast. A recurrent problem is to obtain accurate traffic matrices in real time from the traffic of a network, specially when this network is large (e.g., a Tier 1 Internet Service Provider), and without causing a relevant overhead in network computing, storage and communication resources. A solution deeply investigated in the past is the network tomography: the estimation of a traffic matrix from the volume of traffic traversing the links (a.k.a. link counts), which measurement implies a minimum overhead. This estimation entails relevant challenges. In this paper, we propose the application of the Partial Least Squares method to this problem. We illustrate the proposal with the Abilene network dataset, and report promising results in comparison to traditional methods like General Tomogravity and the Structural Analysis based on Principal Component Analysis.

show abstract