Lossy Compression of Network Feature Data: When Less Is Enough

TL;DR

This paper explores task-aware lossy compression methods for network traffic features, demonstrating that simple, semantics-preserving techniques can significantly reduce storage needs while maintaining analytical accuracy in diverse network environments.

Contribution

It introduces simple, semantics-preserving lossy compression strategies tailored for network traffic features, balancing storage efficiency and analytics performance.

Findings

Stable operating regions identified for compression

Compression maintains accuracy in website classification

Reduces storage without sacrificing task performance

Abstract

Network traffic analysis increasingly relies on feature-based representations to support monitoring and security in the presence of pervasive encryption. Although features are more compact than raw packet traces, their storage has become a scalability bottleneck from large-scale core networks to resource-constrained Internet of Things (IoT) environments. This article investigates task-aware lossy compression strategies that reduce the storage footprint of traffic features while preserving analytics accuracy. Using website classification in core networks and device identification in IoT environments as representative use cases, we show that simple, semantics-preserving compression techniques expose stable operating regions that balance storage efficiency and task performance. These results highlight compression as a first-class design dimension in scalable network monitoring systems.