Spatiotemporal traffic matrix prediction: A deep learning approach with wavelet multiscale analysis

Zhao, Jing; Qu, Hua; Zhao, Jihong; Jiang, Dingchao

doi:10.1002/ett.3640

Cited by 22 publications

(13 citation statements)

References 31 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Since TM estimation is also a similar problem, neural networks have also been used in this area. In [41], [42] the authors also proposed to use neural networks for TM completion.…”

Section: Related Workmentioning

confidence: 99%

Learning Based Methods for Traffic Matrix Estimation From Link Measurements

Kodialam

Lakshman

et al. 2021

IEEE Open J. Commun. Soc.

View full text Add to dashboard Cite

Network traffic matrix (TM) is a critical input for capacity planning, anomaly detection and many other network management related tasks. The TMs are often computed from link load measurements. The TM estimation problem is the determination of the TM from link load measurements. The relationship between the link loads and the TM that generated the link loads can be modeled as an under-determined linear system and has multiple feasible solutions. Therefore, prior knowledge of the traffic demand pattern has to be used in order to find a potentially feasible TM. In this paper, we consider the TM estimation problem with limited prior information. Unlike previous methods that require past measurements of complete TMs, which are hard to obtain or protected by regulations, our method works even if only the distribution of TMs is known. We develop an iterative projection based algorithm to solve this problem. If large number of past TMs can be measured, we propose a Generative Adversarial Network (GAN) based approach for solving the problem. We compare the strengths of the two approaches and evaluate their performance for several networks using varying amounts of past data.

show abstract

“…Since TM estimation is also a similar problem, neural networks have also been used in this area. In [41], [42] the authors also proposed to use neural networks for TM completion.…”

Section: Related Workmentioning

confidence: 99%

Learning Based Methods for Traffic Matrix Estimation From Link Measurements

Kodialam

Lakshman

et al. 2021

IEEE Open J. Commun. Soc.

View full text Add to dashboard Cite

show abstract

“…Vinayakumar et al 24 applied and evaluated various RNN including GRU network in traffic prediction on the GÉANT backbone networks. In addition, in order to capture spatial relationships presented in the traffic networks, many forecasting models 17,25,26 incorporated CNNs to extract spatial features from 2D spatial‐temporal traffic data. Typically, Cao et al 17 proposed a hybrid CNN‐LSTM model which combined CNN and LSTM to discover spatial and temporal features in the data center network simultaneously.…”

Section: Related Workmentioning

confidence: 99%

“…Recently, more advanced and powerful deep learning models have significantly achieved performance gains in traffic prediction 16‐26 . The deep architecture based on deep belief networks (DBN) has been successfully applied to solve TM prediction in large‐scale IP backbone networks 18 .…”

Section: Introductionmentioning

confidence: 99%

“…In general, the transmission of network traffic is built on top of the network topology, so the traffic patterns between adjacent nodes are potentially correlated. To this end, several works 17,25,26 leveraged a combination of convolution neural network (CNN) and RNN to exploit such spatial and temporal regularities. Although CNN‐based model can explicitly mine spatial correlations of traffic flows, the conventional convolution operation applied restricts the model to processing only grid or image data rather than general graph domains.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Spatiotemporal graph convolutional recurrent networks for traffic matrix prediction

Zhao

et al. 2020

Trans Emerging Tel Tech

Self Cite

View full text Add to dashboard Cite

Summary Through accurate network‐wide traffic prediction, network operators can agilely manage resources and improve robustness by proactively adapting to new traffic patterns, especially for traffic engineering, capacity planning and quality of service provisioning. However, due to the proliferation of backbone network traffic as well as the complexity and dynamics of network communication behavior, accurate and effective network‐wide traffic prediction is challenging. To address the challenges, this paper focuses on short‐term traffic matrix (TM) prediction in large‐scale IP backbone networks. In order to improve the prediction performance, a novel spatiotemporal graph convolutional recurrent network (SGCRN)—a deep learning framework that incorporates both spatial and temporal dependencies of traffic flows, is proposed to implement TM prediction with high accuracy and efficiency. By learning network‐wide traffic as graph‐structured TM time series, SGCRN jointly utilizes graph convolutional networks (GCN) and gated recurrent units (GRU) networks to extract comprehensive spatiotemporal correlations among traffic flows. Specifically, SGCRN employs GCNs to identify structural spatial features of traffic flows by considering topological properties, and utilizes GRUs to implement temporal features learning by considering short and long‐term dynamics of traffic flows. Extensive experimental results on the inter‐Points of Presence network traffic data from four real IP backbone networks show that SGCRN can effectively predict short‐term network‐wide TM with superior accuracy compared with other four widely used traffic prediction methods.

show abstract

“…Deep learning method is used to investigate medical injury 17 . This method is also used in different other fields like IoT, and its performance is better than traditional techniques 18‐21 …”

Section: Introductionmentioning

confidence: 99%

Deep learning recognition of diseased and normal cell representation

Iqbal

Ahmad

Luo

et al. 2020

Trans Emerging Tel Tech

View full text Add to dashboard Cite

Cell classification refers to detecting normal and diseased cells from small amount of data. Sometimes, classification of cells becomes difficult because some cells fall into more than one categories/classes. Current state‐of‐the‐art cell classification methods have been developed on the bases of tumor cell classification but these methods cannot classify diseased or normal cells. This study investigated the performance of two classification methods traditional machine learning and deep learning (normal and diseased cell classification) to categorize normal and diseased cells. Millions of normal cells undergo controlled growth and uncontrolled growth may be involved in disease causation but their clinical applications remain limited due to difficulties in distinguishing normal and diseased cells. Previous studies are limited to identify, systematically, the normal and diseased cells. This study collected information about diseased or normal cells to check the networks for correct cell detection and then eliminated false‐positive cells. We used machine learning methods along with logistic regression, support vector machine, and CNN (convolutional neural network). We found that our proposed method classified better the normal and diseased cells. With the help of two types of images: normal and diseased cells, we trained a CNN that identified diseased cells with 98% accuracy and enabled the discovery of normal and diseased cells. As a result, it will advance the clinical utility of human diseased cells.

show abstract

Spatiotemporal traffic matrix prediction: A deep learning approach with wavelet multiscale analysis

Cited by 22 publications

References 31 publications

Learning Based Methods for Traffic Matrix Estimation From Link Measurements

Learning Based Methods for Traffic Matrix Estimation From Link Measurements

Spatiotemporal graph convolutional recurrent networks for traffic matrix prediction

Deep learning recognition of diseased and normal cell representation

Contact Info

Product

Resources

About