Atom: Neural Traffic Compression with Spatio-Temporal Graph Neural Networks
Open Access
- 5 December 2023
- conference paper
- Published by Association for Computing Machinery (ACM)
- p. 1-6
Abstract
Storing network traffic data is key to efficient network management; however, it is becoming more challenging and costly due to the ever-increasing data transmission rates, traffic volumes, and connected devices. In this paper, we explore the use of neural architectures for network traffic compression. Specifically, we consider a network scenario with multiple measurement points in a network topology. Such measurements can be interpreted as multiple time series that exhibit spatial and temporal correlations induced by network topology, routing, or user behavior. We present Atom, a neural traffic compression method that leverages spatial and temporal correlations present in network traffic. Atom implements a customized spatio-temporal graph neural network design that effectively exploits both types of correlations simultaneously. The experimental results show that Atom can outperform GZIP's compression ratios by 50%--65% on three real-world networks.Keywords
All Related Versions
This publication has 16 references indexed in Scilit:
- Inside the Social Network's (Datacenter) NetworkACM SIGCOMM Computer Communication Review, 2015
- Packet-Level Network Compression: Realization and Scaling of the Network-Wide BenefitsIEEE/ACM Transactions on Networking, 2015
- Network traffic characteristics of data centers in the wild2010
- NET-FLiProceedings of the VLDB Endowment, 2010
- High Speed Lossless Data Compression Architecture2006
- A measurement study of correlations of Internet flow characteristicsComputer Networks, 2006
- Wavelength converter placement in least-load-routing-based optical networks using genetic algorithmsJournal of Optical Networking, 2004
- Arithmetic coding for data compressionCommunications of the ACM, 1987
- A universal algorithm for sequential data compressionIEEE Transactions on Information Theory, 1977
- A Method for the Construction of Minimum-Redundancy CodesProceedings of the IRE, 1952