GO-GenZip: Goal-Oriented Generative Sampling and Hybrid Compression

Pietro Talli; Qi Liao; Alessandro Lieto; Parijat Bhattacharjee; Federico Chiariotti; Andrea Zanella

cs.LGMar 20, 2026

GO-GenZip: Goal-Oriented Generative Sampling and Hybrid Compression

Authors: Pietro Talli, Qi Liao, Alessandro Lieto, Parijat Bhattacharjee, Federico Chiariotti, Andrea Zanella

Organizations: Dept. of Information Engineering, University of Padova, Italy. · Nokia Bell Labs Stuttgart, Germany. · Nokia Bengaluru, India.

Abstract

Current network data telemetry pipelines consist of massive streams of fine-grained Key Performance Indicators (KPIs) from multiple distributed sources towards central aggregators, making data storage, transmission, and real-time analysis increasingly unsustainable. This work presents a generative AI (GenAI)-driven sampling and hybrid compression framework that redesigns network telemetry from a goal-oriented perspective. Unlike conventional approaches that passively compress fully observed data, our approach jointly optimizes what to observe and how to encode it, guided by the relevance of information to downstream tasks. The framework integrates adaptive sampling policies, using adaptive masking techniques, with generative modeling to identify patterns and preserve critical features across temporal and spatial dimensions. The selectively acquired data are further processed through a hybrid compression scheme that combines traditional lossless coding with GenAI-driven, lossy compression. Experimental results on real network datasets demonstrate over 50 $\%$ reductions in sampling and data transfer costs, while maintaining comparable reconstruction accuracy and goal-oriented analytical fidelity in downstream tasks.

Figures & tables

Panel of Figure 2 — Fig. 2: GO-GenZip sampling and compression architecture.

Panel of Figure 3 — Fig. 3: Comparison between Generative compression and Hybrid compression for different SRs.

Panel of Figure 4 — Fig. 4: Comparison between Generative compression and Hybrid compression for different CRs.

Panel of Figure 5 — Fig. 5: Comparison of adaptive and fixed method with respect to SR and CR.

Panel of Figure 6 — Fig. 6: Visualization of masks for different BS classes and prediction tasks. White, blue, red regions represent the unsampled, compression with generative model, and compression with LZMA , respectively.

Method	Task1	Task2	Task3	Task4
Recon-Based	1.27±0.04	0.09±0.004	0.04±0.003	1.48±0.14
GO E2E	1.21±0.02	0.08±0.003	0.04±0.0008	1.18±0.07

TABLE I: Prediction MAE with and without GO end-to-end training.

Explore similar work

Lightweight GenAI for Network Traffic Generation: Fidelity, Augmentation, and Classification
Mar 26, 2026Giampaolo Bovenzi, Domenico Ciuonzo, Jonatan Krolikowski +4Synthetic Data Generation Synthetic Data Augmentation
DIETI, University of Naples Federico II, Italy. · Huawei Technologies France SASU.
Learning Compression Rules for Network Traffic
Aug 5, 2026Quentin Lampin, Éloi Sainte-Beuve, Louis-Adrien Dufrène +2
Orange Research · 1Orange Research · Université Grenoble Alpes +1
Transforming the Use of Earth Observation Data: Exascale Training of a Generative Compression Model with Historical Priors for up to 10,000x Data Reduction
May 9, 2026Jinxiao Zhang, Runmin Dong, Xiyong Wu +14Learned Image Compression Generative Image Compression
Department of Earth System Science, Tsinghua University · Sun Yat-Sen University · Institute of Data and Information, Tsinghua Shenzhen International Graduate School +2

Mar 26, 2026J/K move · Enter open · S save

Lightweight GenAI for Network Traffic Generation: Fidelity, Augmentation, and Classification

Giampaolo Bovenzi, Domenico Ciuonzo, Jonatan Krolikowski, Antonio Montieri+3

DIETI, University of Naples Federico II, Italy. · Huawei Technologies France SASU.

Synthetic Data Generation Synthetic Data Augmentation

Open paper AskPDF arXiv

Network Traffic Classification (NTC) increasingly relies on data-driven models, yet its practical deployment is often constrained by limited labeled data, strict privacy requirements, and the cost of collecting representative traffic traces. While Network Traffic Generation (NTG) provides an effective means to mitigate data scarcity, conventional generative methods struggle to model the complex temporal dynamics of modern traffic and often incur high computational costs. In this article, we investigate lightweight Generative Artificial Intelligence (GenAI) architectures for practical NTG. Rather than generating raw packet bytes or relying on large foundation models, we synthesize compact flow-level traffic representations derived from early packet-header information, enabling transformer-based, state-space, and diffusion models with only a few million parameters. We present a modular GenAI pipeline for NTG and evaluate it along four complementary axes: (i) synthetic traffic fidelity, (ii) synthetic-only training for privacy-preserving NTC, (iii) data augmentation under low-data regimes, and (iv) computational efficiency. Experiments on two heterogeneous datasets show that lightweight transformer-based and state-space models preserve both static and temporal traffic characteristics, while providing useful synthetic data for downstream NTC. Among them, transformer-based models offer the best fidelity-efficiency trade-off, combining high-quality traffic generation with moderate computational overhead.

Abstract

Explore similar work

Lightweight GenAI for Network Traffic Generation: Fidelity, Augmentation, and Classification

Learning Compression Rules for Network Traffic

Transforming the Use of Earth Observation Data: Exascale Training of a Generative Compression Model with Historical Priors for up to 10,000x Data Reduction