# network-intelligence

Network intelligence is a technology that analyzes network traffic, including packet content and metadata, in real time for network management and cybersecurity.

This is one page of public article previews, not the complete archive. Follow Next page to continue. Summaries are not the original full articles.

## Do LLMs Understand Network Traces? A Comparative Study of Reasoning over Raw and Filtered Traffic Data

DevFeed: [Do LLMs Understand Network Traces? A Comparative Study of Reasoning over Raw and Filtered Traffic Data](<https://devfeed.tech/articles/do-llms-understand-network-traces-a-comparative-study-of-reasoning-over-raw-and-filtered-traffic-data-140051.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2611.22575>)

Author: Irene Gos\\'alvez-White, Roberto Mag\\'an-Carri\\'on, Carlota de la Vega-Soriano, Pedro Garc\\'ia-Teodoro

Published: 2026-11-18T05:00:00Z

Content type: comparison

Language: en

Sources: [arXiv: Artificial Intelligence](<https://devfeed.tech/sources/arxiv-artificial-intelligence.md>), [arXiv: Networking](<https://devfeed.tech/sources/arxiv-networking.md>)

Topics: [network-intelligence](<https://devfeed.tech/topics/network-intelligence.md>), [LLM evaluation / benchmarking](<https://devfeed.tech/topics/llm-evaluation-benchmarking.md>), [Large Language Model](<https://devfeed.tech/topics/llm.md>), [Internet Traffic](<https://devfeed.tech/topics/internet-traffic.md>), [llm-reasoning](<https://devfeed.tech/topics/llm-reasoning.md>), [Ground truth / benchmark quality](<https://devfeed.tech/topics/ground-truth-benchmark-quality.md>)

Tags: [cs-ai](<https://devfeed.tech/tags/cs-ai.md>), [cs-ni](<https://devfeed.tech/tags/cs-ni.md>), [dataset](<https://devfeed.tech/tags/dataset.md>), [evaluation](<https://devfeed.tech/tags/evaluation.md>), [llms](<https://devfeed.tech/tags/llms.md>), [network-traffic-analysis](<https://devfeed.tech/tags/network-traffic-analysis.md>), [optimization](<https://devfeed.tech/tags/optimization.md>), [overhead](<https://devfeed.tech/tags/overhead.md>), [packet](<https://devfeed.tech/tags/packet.md>), [performance](<https://devfeed.tech/tags/performance.md>), [protocol](<https://devfeed.tech/tags/protocol.md>), [reliability](<https://devfeed.tech/tags/reliability.md>)

### AI overview

Using benign CICIDS2017 traffic and human-generated ground truth, the study evaluates multiple LLMs on packet-field extraction and flow-sequence reconstruction from raw traffic dumps versus protocol-filtered representations. Responses are scored with an automated LLM-as-a-judge pipeline validated against human evaluations. The abstract reports that protocol filtering reduces prompt overhead and improves answer accuracy by reducing incorrect responses, suggesting it can make LLM-assisted network analysis more efficient and reliable. The excerpt does not provide model-specific results or quantitative measurements.

### Source excerpt

arXiv:2611.22575v1 Announce Type: cross Abstract: The effective delivery and monitoring of ICT services rely heavily on network traffic analysis. However, interpreting complex packet captures remains a challenging task that requires substantial technical expertise. Large language models (LLMs) offer promising opportunities to automate network traffic in- vestigation, yet their reliability in packet-level field extraction and flow-level sequence reconstruction remains an open research question. This study proposes a reproducible, question-answering- based evaluation methodology to assess the capabilities of LLMs in network trace analysis. Using real-world benign traffic from the CICIDS2017 dataset, we evaluate multiple models against a human-generated ground truth, comparing their performance on raw textual traffic dumps and protocol-filtered representations. Model responses are assessed through an automated LLM- as-a-judge pipeline, validated against human evaluations. The results demonstrate that protocol filtering significantly reduces prompt overhead while improving answer accuracy by reducing incorrect responses. These findings highlight the importance of protocol-aware traffic filtering as an effective optimization for improving the efficiency and reliability of LLM-assisted network management.

## HiGAI-Net: Hierarchical Encrypted Traffic Classification for Generative AI Services

DevFeed: [HiGAI-Net: Hierarchical Encrypted Traffic Classification for Generative AI Services](<https://devfeed.tech/articles/higai-net-hierarchical-encrypted-traffic-classification-for-generative-ai-services-140401.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2611.22373>)

Author: Madhan Raj Kanagarathinam, Jaheon Gu, Krishna M. Sivalingam, Kapal Dev

Published: 2026-11-18T05:00:00Z

Content type: article

Language: en

Sources: [arXiv: Networking](<https://devfeed.tech/sources/arxiv-networking.md>)

Topics: [network-intelligence](<https://devfeed.tech/topics/network-intelligence.md>), [Internet Traffic](<https://devfeed.tech/topics/internet-traffic.md>)

Tags: [ai](<https://devfeed.tech/tags/ai.md>), [ai-gate](<https://devfeed.tech/tags/ai-gate.md>), [ai-services](<https://devfeed.tech/tags/ai-services.md>), [chatgpt](<https://devfeed.tech/tags/chatgpt.md>), [classification](<https://devfeed.tech/tags/classification.md>), [classifier](<https://devfeed.tech/tags/classifier.md>), [copilot](<https://devfeed.tech/tags/copilot.md>), [cs-ni](<https://devfeed.tech/tags/cs-ni.md>), [deep-learning](<https://devfeed.tech/tags/deep-learning.md>), [encrypted](<https://devfeed.tech/tags/encrypted.md>), [encrypted-traffic](<https://devfeed.tech/tags/encrypted-traffic.md>), [flow](<https://devfeed.tech/tags/flow.md>), [gemini](<https://devfeed.tech/tags/gemini.md>), [genai](<https://devfeed.tech/tags/genai.md>), [generative-ai](<https://devfeed.tech/tags/generative-ai.md>), [gradient](<https://devfeed.tech/tags/gradient.md>), [metadata](<https://devfeed.tech/tags/metadata.md>), [qos](<https://devfeed.tech/tags/qos.md>), [quic](<https://devfeed.tech/tags/quic.md>), [radio](<https://devfeed.tech/tags/radio.md>), [traffic](<https://devfeed.tech/tags/traffic.md>)

### AI overview

HiGAI-Net identifies generative AI traffic from packet-header metadata despite limited visibility into encrypted handshakes. It stacks three classifiers: an AI/non-AI gate (98.5% weighted F1), a resolver for ChatGPT, Copilot, and Gemini (86.4%), and a text-versus-voice/video classifier (61.0%). Its sliding-window features track changes in download/upload byte ratios and streaming regularity. On 44,492 labeled flows, gradient-boosted trees outperformed the tested deep-learning variants; caching per-flow labels reduced amortized inference to below 1 μs per packet. These results describe the reported dataset and evaluation.

### Source excerpt

arXiv:2611.22373v1 Announce Type: new Abstract: The Radio Access Network (RAN) today does not have a reliable way to identify GenAI traffic (ChatGPT, Microsoft Copilot, or Google Gemini traffic) at the bearer level. TLS~1.3 moves SNI into the encrypted handshake record, and Encrypted ClientHello (ECH) removes it entirely from passive inspection; Gemini has partly shifted to QUIC where SNI visibility also depends on whether ECH is deployed. Without per-flow application identity, differentiated QoS for GenAI services is not possible. HiGAI-Net tackles this using only packet-header metadata, stacking three independently trained classifiers: a binary AI/Non-AI gate (98.5 % weighted F1), a per-service resolver across ChatGPT, Copilot, and Gemini (86.4 %), and a text-vs-voice/video mode classifier (61.0 %). The core innovation is a suite of sliding-window temporal features that track how DL/UL byte-ratios evolve within a session -- the per-quarter and per-half ratio trends, a trend differential $\Delta R$, and a streaming regularity index. Tested on 44,492 labeled flows, gradient-boosted tree models consistently outperformed all deep-learning variants; a per-flow label cache brings amortised inference below 1~$\mu$s per packet.

## Autonomous Cyber Operations Gym for Training and Evaluating AI Models for Defence

DevFeed: [Autonomous Cyber Operations Gym for Training and Evaluating AI Models for Defence](<https://devfeed.tech/articles/autonomous-cyber-operations-gym-for-training-and-evaluating-ai-models-for-defence-133267.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2610.18594>)

Author: Tobias H\\"urten, Johannes F. Loevenich, Florian Spelter, Erik Adler, Laurin Holz, Maxime Schwarzer, Thies Moehlenhof, Roberto Rigolin F. Lopes

Published: 2026-11-13T05:00:00Z

Content type: article

Language: en

Sources: [arXiv: Cryptography and Security](<https://devfeed.tech/sources/arxiv-cryptography-and-security.md>)

Topics: [Detection engineering](<https://devfeed.tech/topics/detection-engineering.md>), [Artificial Intelligence](<https://devfeed.tech/topics/ai.md>), [network-intelligence](<https://devfeed.tech/topics/network-intelligence.md>), [netflow](<https://devfeed.tech/topics/netflow.md>)

Tags: [abstract](<https://devfeed.tech/tags/abstract.md>), [adversarial](<https://devfeed.tech/tags/adversarial.md>), [agents](<https://devfeed.tech/tags/agents.md>), [ai](<https://devfeed.tech/tags/ai.md>), [ai-models](<https://devfeed.tech/tags/ai-models.md>), [arxiv](<https://devfeed.tech/tags/arxiv.md>), [autonomous](<https://devfeed.tech/tags/autonomous.md>), [based](<https://devfeed.tech/tags/based.md>), [behaviour](<https://devfeed.tech/tags/behaviour.md>), [can](<https://devfeed.tech/tags/can.md>), [changes](<https://devfeed.tech/tags/changes.md>), [components](<https://devfeed.tech/tags/components.md>), [composition](<https://devfeed.tech/tags/composition.md>), [cs-cr](<https://devfeed.tech/tags/cs-cr.md>), [cs-lg](<https://devfeed.tech/tags/cs-lg.md>), [cyber](<https://devfeed.tech/tags/cyber.md>), [cybersecurity](<https://devfeed.tech/tags/cybersecurity.md>), [datasets](<https://devfeed.tech/tags/datasets.md>), [gym](<https://devfeed.tech/tags/gym.md>), [models](<https://devfeed.tech/tags/models.md>), [operations](<https://devfeed.tech/tags/operations.md>), [training](<https://devfeed.tech/tags/training.md>)

### AI overview

The Autonomous Cyber Operations gym combines high-fidelity network emulation, models of benign user behavior and adversarial profiles, and open toolkits for generating cybersecurity scenarios. It supports live or deferred packet- and flow-level feature extraction to create labeled synthetic data for training DNN intrusion detectors, and emulated interactions for developing and evaluating autonomous defence components. A DNN classifier achieved near-perfect recall on adversarial flows in-domain, but even small changes in network composition, including protocol distribution, caused a reported collapse in transferability and detection performance.

### Source excerpt

arXiv:2610.18594v2 Announce Type: replace Abstract: This paper proposes an Autonomous Cyber Operations (ACO) gym, a unified framework integrating high-fidelity network emulation, realistic user behaviour modelling for benign network traffic and adversarial profiles, and the generation of cybersecurity scenarios through open toolkits. By supporting live and deferred feature extraction at the packet and flow levels, the gym facilitates the creation of labelled synthetic datasets for training Deep Neural Network (DNN)-based Intrusion Detection Systems (IDSs), as well as developing and evaluating modular components for autonomous defence agents through emulated interactions. While evaluation of a DNN classifier within the gym shows high in-domain performance, achieving near-perfect recall on adversarial network flows, our experiments reveal a significant sensitivity to domain shift, showing that even minor changes in network composition, such as variations in traffic protocol distribution, can lead to a catastrophic collapse in model transferability and detection performance.

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

DevFeed: [Lightweight GenAI for Network Traffic Generation: Fidelity, Augmentation, and Classification](<https://devfeed.tech/articles/lightweight-genai-for-network-traffic-generation-fidelity-augmentation-and-classification-131571.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2603.25507>)

Author: Giampaolo Bovenzi, Domenico Ciuonzo, Jonatan Krolikowski, Antonio Montieri, Alfredo Nascita, Antonio Pescap\\\`e, Dario Rossi

Published: 2026-11-13T05:00:00Z

Content type: article

Language: en

Sources: [arXiv: Artificial Intelligence](<https://devfeed.tech/sources/arxiv-artificial-intelligence.md>), [arXiv: Networking](<https://devfeed.tech/sources/arxiv-networking.md>)

Topics: [network-intelligence](<https://devfeed.tech/topics/network-intelligence.md>), [Internet Traffic](<https://devfeed.tech/topics/internet-traffic.md>), [data augmentation](<https://devfeed.tech/topics/data-augmentation.md>)

Tags: [classification](<https://devfeed.tech/tags/classification.md>), [cs-ai](<https://devfeed.tech/tags/cs-ai.md>), [cs-lg](<https://devfeed.tech/tags/cs-lg.md>), [cs-ni](<https://devfeed.tech/tags/cs-ni.md>), [data-augmentation](<https://devfeed.tech/tags/data-augmentation.md>), [diffusion](<https://devfeed.tech/tags/diffusion.md>), [efficiency](<https://devfeed.tech/tags/efficiency.md>), [genai](<https://devfeed.tech/tags/genai.md>), [generation](<https://devfeed.tech/tags/generation.md>), [lightweight](<https://devfeed.tech/tags/lightweight.md>), [network-traffic](<https://devfeed.tech/tags/network-traffic.md>), [state-space-models](<https://devfeed.tech/tags/state-space-models.md>), [synthetic-data](<https://devfeed.tech/tags/synthetic-data.md>), [transformer](<https://devfeed.tech/tags/transformer.md>)

### AI overview

The study evaluates lightweight generative models for network traffic generation using compact flow representations derived from early packet headers, rather than raw packet bytes or large foundation models. Models with a few million parameters are assessed for traffic fidelity, downstream classification trained only on synthetic data, low-data augmentation, and computational efficiency. On two heterogeneous datasets, transformer and state-space models preserve static and temporal traffic characteristics and produce useful data for network traffic classification. Transformer models provide the best reported fidelity-efficiency trade-off, with high-quality generation and moderate computational overhead.

### Source excerpt

arXiv:2603.25507v3 Announce Type: replace Abstract: 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 reduced raw-data sharing in 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.

## Versatile yet Efficient Network Traffic Analysis: Offloading Network Foundation Model to SmartNIC

DevFeed: [Versatile yet Efficient Network Traffic Analysis: Offloading Network Foundation Model to SmartNIC](<https://devfeed.tech/articles/versatile-yet-efficient-network-traffic-analysis-offloading-network-foundation-model-to-smartnic-131569.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2508.02001>)

Author: Chungang Lin, Xuying Meng, Tianyu Zuo, Weiyao Zhang, Meng Shen, Ruijie Zhao, Guanming Che, Ruiqi Meng, Ziyue Huang, Haitong Luo, Zhiwei Xu, Yujun Zhang

Published: 2026-11-13T05:00:00Z

Content type: article

Language: en

Sources: [arXiv: Networking](<https://devfeed.tech/sources/arxiv-networking.md>)

Topics: [network-intelligence](<https://devfeed.tech/topics/network-intelligence.md>), [Networking hardware](<https://devfeed.tech/topics/networking-hardware.md>), [model architecture](<https://devfeed.tech/topics/model-architecture.md>)

Tags: [analysis](<https://devfeed.tech/tags/analysis.md>), [architecture](<https://devfeed.tech/tags/architecture.md>), [convolution](<https://devfeed.tech/tags/convolution.md>), [cs-lg](<https://devfeed.tech/tags/cs-lg.md>), [cs-ni](<https://devfeed.tech/tags/cs-ni.md>), [edge](<https://devfeed.tech/tags/edge.md>), [efficiency](<https://devfeed.tech/tags/efficiency.md>), [encoding](<https://devfeed.tech/tags/encoding.md>), [encryption](<https://devfeed.tech/tags/encryption.md>), [execution](<https://devfeed.tech/tags/execution.md>), [foundation](<https://devfeed.tech/tags/foundation.md>), [foundation-model](<https://devfeed.tech/tags/foundation-model.md>), [latency](<https://devfeed.tech/tags/latency.md>), [model](<https://devfeed.tech/tags/model.md>), [model-architecture](<https://devfeed.tech/tags/model-architecture.md>), [modeling](<https://devfeed.tech/tags/modeling.md>), [network](<https://devfeed.tech/tags/network.md>), [network-traffic](<https://devfeed.tech/tags/network-traffic.md>), [network-traffic-analysis](<https://devfeed.tech/tags/network-traffic-analysis.md>), [offloading](<https://devfeed.tech/tags/offloading.md>), [traffic](<https://devfeed.tech/tags/traffic.md>), [versatile](<https://devfeed.tech/tags/versatile.md>)

### AI overview

Nepco aims to combine flexible traffic analysis with low-latency edge execution by offloading a network foundation model to an NVIDIA BlueField-3 SmartNIC. It directly embeds raw byte sequences to avoid complex encoding, then uses an observation-driven convolution architecture with hardware-friendly scoring and gating to locate salient patterns despite shifts in byte offsets. Multi-engine collaborative execution supports the prototype. The excerpt reports macro F1 competitive with the best results of eight state-of-the-art network foundation models; it does not provide specific scores or further evaluation details.

### Source excerpt

arXiv:2508.02001v3 Announce Type: replace Abstract: Pervasive encryption makes large-scale labeling infeasible for traffic analysis, while security operations demand edge analysis to avert service degradation and vulnerabilities. These pressures have produced two research lines that advance 1) versatile analysis, via network foundation models for low label dependency, and 2) efficient analysis, via hardware offloading for low analysis latency. However, co-achieving versatility and efficiency via such lines has appeared incompatible, with prior work advancing one while compromising the other, yet we show that this incompatibility is a consequence of polarized design choices across the three components of traffic analysis systems, i.e., traffic processing, model architecture, and analysis execution. In response, we present Nepco, a versatile yet efficient network traffic analysis system that offloads network foundation models to SmartNIC. Our key observation is that discriminative traffic information often exhibits localized patterns within byte sequences, motivating versatile yet efficient localized byte-sequence modeling. To exploit this without incurring the latency bottlenecks of complex encoding steps, we employ a hardware-friendly processing strategy that directly embeds raw byte sequences. To maintain versatility across diverse tasks characterized by spatially shifting features, we explore an observation-driven convolution architecture equipped with hardware-friendly scoring and gating designs. By exploiting translation invariance, this design dynamically locates and extracts salient semantic signatures irrespective of their byte-level offsets. We prototype Nepco on the NVIDIA BlueField-3 SmartNIC with multi-engine collaborative analysis execution. The experimental results demonstrate that Nepco achieves macro F1 competitive with the best performances achieved by 8 state-of-the-art network foundation models...

## Theory-Driven Network Backbones for Bipartite Projections

DevFeed: [Theory-Driven Network Backbones for Bipartite Projections](<https://devfeed.tech/articles/theory-driven-network-backbones-for-bipartite-projections-132703.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2611.16692>)

Author: Nikos Salamanos, Paula Gheorghiade, Tom Brughmans, Michele Coscia

Published: 2026-11-13T05:00:00Z

Content type: article

Language: en

Sources: [arXiv: Databases](<https://devfeed.tech/sources/arxiv-databases.md>)

Topics: [network-intelligence](<https://devfeed.tech/topics/network-intelligence.md>)

Tags: [algorithms](<https://devfeed.tech/tags/algorithms.md>), [cs-db](<https://devfeed.tech/tags/cs-db.md>), [cs-si](<https://devfeed.tech/tags/cs-si.md>), [experiments](<https://devfeed.tech/tags/experiments.md>), [inference](<https://devfeed.tech/tags/inference.md>), [model](<https://devfeed.tech/tags/model.md>), [network](<https://devfeed.tech/tags/network.md>), [nodes](<https://devfeed.tech/tags/nodes.md>), [recovery](<https://devfeed.tech/tags/recovery.md>), [relationships](<https://devfeed.tech/tags/relationships.md>), [simulation](<https://devfeed.tech/tags/simulation.md>), [sites](<https://devfeed.tech/tags/sites.md>), [structure](<https://devfeed.tech/tags/structure.md>), [study](<https://devfeed.tech/tags/study.md>), [synthetic](<https://devfeed.tech/tags/synthetic.md>), [test](<https://devfeed.tech/tags/test.md>), [theory](<https://devfeed.tech/tags/theory.md>)

### AI overview

The paper introduces theory-driven backboning for bipartite networks, which uses a user-specified Agent-Based Model to represent how two types of agents generate links. Simulation-Based Inference fits that model to the observed network and reconstructs likely true relationships, incorporating domain dynamics rather than relying on data-agnostic edge-filtering tests used by structural and statistical backboning. In synthetic experiments, the approach improves latent-edge recovery over conventional methods. An archaeological case study of amphorae from different sites yields interpretable patterns despite imperfect model fit. The evidence here is the abstract, so it does not establish details about implementation or performance beyond these reported findings.

### Source excerpt

arXiv:2611.16692v1 Announce Type: cross Abstract: Collecting data, including network data, is subject to measurement error. Network backboning is the process of statistically validating each edge in a network to estimate the probability that the edge is born of noisy measurements. The current backboning literature provides approaches in two classes: structural and statistical. The classes differ in many respects, but they share one commonality: the test used to determine whether to filter out noisy edges is data-agnostic; it does not take into account what the network is representing. However, the nodes and edges of all non-synthetic networks correspond to specific real-world entities, whose dynamics determine how the edge-generating process operates. In this paper, we propose the first in a new class of network backboning algorithms: a theory-driven backboning approach that allows us to encode known dynamics into the backboning process. The resulting backbone embeds the encoded theory and produces a network compatible with it, rather than one that obeys a data-agnostic network process, as in the structural and statistical backboning classes. We specifically focus on the case of projecting and backboning a bipartite structure. We do so by allowing users to specify an Agent-Based Model for the two types of agents (nodes) in the network and by performing Simulation-Based Inference to fit the observed bipartite structure and, as a result, reconstruct the true relationships between the two classes of agents. We test our framework both on synthetic data -- where we have perfect information about the edge-generating process -- and in a case study using archaeological data on amphorae retrieved from different archaeological sites. Our approach improves latent-edge recovery over conventional backboning methods in synthetic experiments and reveals interpretable archaeological patterns despite imperfect model fit.

## Getting Traffic From aocr.org? It Could Be a ByteDance Crawler

DevFeed: [Getting Traffic From aocr.org? It Could Be a ByteDance Crawler](<https://devfeed.tech/articles/getting-traffic-from-aocr-org-it-could-be-a-bytedance-crawler-130186.md>)

Original publisher: [Read original article](<https://hackernoon.com/getting-traffic-from-aocrorg-it-could-be-a-bytedance-crawler?source=rss>)

Author: Nandan Kumar

Published: 2026-11-12T06:59:11Z

Content type: article

Language: en

Sources: [HackerNoon](<https://devfeed.tech/sources/hackernoon.md>)

Topics: [Internet Traffic](<https://devfeed.tech/topics/internet-traffic.md>), [network security](<https://devfeed.tech/topics/network-security.md>), [network-intelligence](<https://devfeed.tech/topics/network-intelligence.md>), [Google Analytics](<https://devfeed.tech/topics/google-analytics.md>), [residential proxy network](<https://devfeed.tech/topics/residential-proxy-network.md>)

Tags: [analytics](<https://devfeed.tech/tags/analytics.md>), [aocr-org-referral-spam](<https://devfeed.tech/tags/aocr-org-referral-spam.md>), [blocking](<https://devfeed.tech/tags/blocking.md>), [browser](<https://devfeed.tech/tags/browser.md>), [bytedance](<https://devfeed.tech/tags/bytedance.md>), [cloud-computing](<https://devfeed.tech/tags/cloud-computing.md>), [cloudflare](<https://devfeed.tech/tags/cloudflare.md>), [crawler](<https://devfeed.tech/tags/crawler.md>), [cybersecurity](<https://devfeed.tech/tags/cybersecurity.md>), [data-analytics](<https://devfeed.tech/tags/data-analytics.md>), [data-science](<https://devfeed.tech/tags/data-science.md>), [git](<https://devfeed.tech/tags/git.md>), [mobile-development](<https://devfeed.tech/tags/mobile-development.md>), [traffic](<https://devfeed.tech/tags/traffic.md>), [website-crawler](<https://devfeed.tech/tags/website-crawler.md>)

### AI overview

The aocr.org referral is forged and appears to expose crawler traffic; the evidence points to a likely ByteDance crawler stack, but does not prove who operates it. Across 199 of 199 examined requests, TLS, HTTP/2, and browser-timing fingerprints matched Bytespider, while real-visitor recordings lacked the unusual timing fields. Requests also showed a Windows disguise, China-based servers behind rotating US proxies, and near-total bounce behavior. The article advises against blocking IPs because they change per visit or contacting AOCR, which is not implicated. It recommends starting a Cloudflare Managed Challenge rule in Log mode using the Accept header and missing sec-ch-ua signal, adding a separate aocr.org referrer rule, and filtering analytics by referrer; it identifies Asia/Shanghai timezone plus 800-pixel screen width as a more durable filter if the referrer changes.

### Source excerpt

Fake aocr.org referrals are appearing across websites. Learn what the evidence reveals about a likely ByteDance crawler stack and how to block it.

## Citrus Ransomware: VLESS for traffic, Sn00py for encryption, and LLMs lending a hand

DevFeed: [Citrus Ransomware: VLESS for traffic, Sn00py for encryption, and LLMs lending a hand](<https://devfeed.tech/articles/citrus-ransomware-vless-for-traffic-sn00py-for-encryption-and-llms-lending-a-hand-129088.md>)

Original publisher: [Read original article](<https://securelist.com/citrus-ransomware/121879/>)

Author: Stanislav Larinsky, Vladislav Vlasov, Boris Trefilov

Published: 2026-11-11T10:00:25Z

Content type: article

Language: en

Sources: [Securelist](<https://devfeed.tech/sources/securelist.md>)

Topics: [ransomware](<https://devfeed.tech/topics/ransomware.md>), [Large Language Model](<https://devfeed.tech/topics/llm.md>), [network-intelligence](<https://devfeed.tech/topics/network-intelligence.md>), [Cryptography](<https://devfeed.tech/topics/cryptography.md>)

Tags: [attacks](<https://devfeed.tech/tags/attacks.md>), [citrus-ransomware](<https://devfeed.tech/tags/citrus-ransomware.md>), [crimeware-reports](<https://devfeed.tech/tags/crimeware-reports.md>), [encryption](<https://devfeed.tech/tags/encryption.md>), [full](<https://devfeed.tech/tags/full.md>), [group-policies](<https://devfeed.tech/tags/group-policies.md>), [large](<https://devfeed.tech/tags/large.md>), [links](<https://devfeed.tech/tags/links.md>), [llm](<https://devfeed.tech/tags/llm.md>), [llms](<https://devfeed.tech/tags/llms.md>), [machine-learning](<https://devfeed.tech/tags/machine-learning.md>), [malware](<https://devfeed.tech/tags/malware.md>), [medium](<https://devfeed.tech/tags/medium.md>), [powershell](<https://devfeed.tech/tags/powershell.md>), [ransomware](<https://devfeed.tech/tags/ransomware.md>), [sn00py](<https://devfeed.tech/tags/sn00py.md>), [thumbnail](<https://devfeed.tech/tags/thumbnail.md>), [tooling](<https://devfeed.tech/tags/tooling.md>), [toy-ghouls](<https://devfeed.tech/tags/toy-ghouls.md>), [traffic](<https://devfeed.tech/tags/traffic.md>), [traffic-tunneling](<https://devfeed.tech/tags/traffic-tunneling.md>), [trojan](<https://devfeed.tech/tags/trojan.md>), [ttps](<https://devfeed.tech/tags/ttps.md>), [vulnerabilities](<https://devfeed.tech/tags/vulnerabilities.md>), [windows-malware](<https://devfeed.tech/tags/windows-malware.md>)

### AI overview

Citrus gains access through compromised contractors' VPN or RDP credentials, then combines familiar intrusion tools with VLESS tunnels and LLM-generated commands and scripts. It uses scheduled tasks, Impacket, credential theft, and tools such as fscan and ADRecon for persistence and discovery; attempts privilege escalation through a certificate authentication bypass and insecure 1C:Enterprise configurations; and moves laterally over SSH, WinRM, and RDP. Xray configures reverse VLESS tunnels on port 443, with REALITY disguising traffic as requests to legitimate sites. The group disables security protections and deploys Sn00py ransomware through malicious Group Policy objects. Sn00py encrypts Windows user files with RSA-2048 and ChaCha20, appends .citrus, and cannot currently be decrypted without the attackers' private key. The article also reports Telegram data collection and similarities to Toy Ghouls, including shared tools and incidents on the same infrastructure, while noting no direct infrastructure overlap.

### Source excerpt

We break down the Citrus Ransomware group's attacks: the attackers' TTPs and tooling, including VLESS traffic tunneling and LLM use, plus their links to Toy Ghouls.

## Verifiable Agentic Traffic Class: A Policed Quality-of-Service Mechanism for Agentic AI Traffic, Evaluated from Simulation to Real-Kernel Adversarial Testing

DevFeed: [Verifiable Agentic Traffic Class: A Policed Quality-of-Service Mechanism for Agentic AI Traffic, Evaluated from Simulation to Real-Kernel Adversarial Testing](<https://devfeed.tech/articles/verifiable-agentic-traffic-class-a-policed-quality-of-service-mechanism-for-agentic-ai-traffic-evaluated-from-simulation-to-real-kernel-adversarial-testing-127848.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2611.14211>)

Author: Naveen Babu Narayanan Nair, Sajay Kurup

Published: 2026-11-11T05:00:00Z

Content type: article

Language: en

Sources: [arXiv: Artificial Intelligence](<https://devfeed.tech/sources/arxiv-artificial-intelligence.md>), [arXiv: Machine Learning](<https://devfeed.tech/sources/arxiv-machine-learning.md>), [arXiv: Networking](<https://devfeed.tech/sources/arxiv-networking.md>)

Topics: [network-intelligence](<https://devfeed.tech/topics/network-intelligence.md>), [LLM security](<https://devfeed.tech/topics/llm-security.md>)

Tags: [adversarial](<https://devfeed.tech/tags/adversarial.md>), [adversarial-testing](<https://devfeed.tech/tags/adversarial-testing.md>), [agent](<https://devfeed.tech/tags/agent.md>), [agent-harnesses](<https://devfeed.tech/tags/agent-harnesses.md>), [agentic](<https://devfeed.tech/tags/agentic.md>), [agentic-ai](<https://devfeed.tech/tags/agentic-ai.md>), [agents](<https://devfeed.tech/tags/agents.md>), [ai](<https://devfeed.tech/tags/ai.md>), [class](<https://devfeed.tech/tags/class.md>), [cs-ai](<https://devfeed.tech/tags/cs-ai.md>), [cs-lg](<https://devfeed.tech/tags/cs-lg.md>), [cs-ni](<https://devfeed.tech/tags/cs-ni.md>), [kernel](<https://devfeed.tech/tags/kernel.md>), [large-language-models](<https://devfeed.tech/tags/large-language-models.md>), [latency](<https://devfeed.tech/tags/latency.md>), [limit](<https://devfeed.tech/tags/limit.md>), [linux](<https://devfeed.tech/tags/linux.md>), [mechanism](<https://devfeed.tech/tags/mechanism.md>), [quality](<https://devfeed.tech/tags/quality.md>), [service](<https://devfeed.tech/tags/service.md>), [simulation](<https://devfeed.tech/tags/simulation.md>), [testing](<https://devfeed.tech/tags/testing.md>), [traffic](<https://devfeed.tech/tags/traffic.md>)

### AI overview

VATC pairs a reserved DSCP marking for agent traffic with router-side classification that demotes flows contradicting the claim, plus a rate-limited priority queue. It matched static DiffServ under steady load; in a simulated burst, assuming an immediate verdict, it reduced mean agent queueing delay by 98% versus DiffServ while increasing human delay by 12%. In a Linux-kernel spoofed flood, its rate limit kept human latency 11 times below DiffServ's. Verification had limits: a live classifier demoted all 20 bulk-shaped spoofers but missed six that copied agent traffic patterns, and emulated network delay erased the separation provided by one timing feature. Without verification, DiffServ and tightly configured fq_codel gave legitimate agents lower latency than VATC. These results used synthetic or sandbox-scale traffic.

### Source excerpt

arXiv:2611.14211v1 Announce Type: cross Abstract: Software agents built on large language models generate traffic unlike human web traffic, yet routers have no class for it and cannot verify a sender's claim to be an agent. This paper proposes the Verifiable Agentic Traffic Class (VATC), which combines a reserved Differentiated Services Code Point (DSCP) value declaring agent traffic, a router-side classifier that demotes flows contradicting the declaration to a class served after best-effort traffic, and a rate-limited priority queue. VATC was compared with static Differentiated Services (DiffServ), a dedicated class that trusts the marking, and fq_codel, a per-flow fair-queuing scheduler, in two simulators and on a real Linux kernel. VATC matched DiffServ under steady load. In a constructed burst it cut mean agent queuing delay by 98% relative to DiffServ in simulation (assuming an immediate verdict), at a 12% increase in human delay; most of the gain came from the separate class. Under a spoofed flood in the kernel, the rate limit held human latency 11 times below DiffServ's. Without verification, DiffServ and a tightly configured fq_codel gave legitimate agents lower latency than VATC; a live four-feature classifier then demoted all 20 bulk-shaped spoofers and gave agents 0.30 ms, but missed six spoofers that copied the agent pattern. With emulated network delay, a timing feature of the classifier stopped separating agent from human flows. In real agent harnesses on a congested link, router queueing took 13% to 63% of task time, depending on harness and assumed model latency. The traffic was synthetic or sandbox-scale.

## Programmable GPU-Accelerated dApps for Real-Time Edge Inference and ISAC on NVIDIA Aerial Testbed

DevFeed: [Programmable GPU-Accelerated dApps for Real-Time Edge Inference and ISAC on NVIDIA Aerial Testbed](<https://devfeed.tech/articles/programmable-gpu-accelerated-dapps-for-real-time-edge-inference-and-isac-on-nvidia-aerial-testbed-127830.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2512.06493>)

Author: Davide Villa, Mauro Belgiovine, Nicholas Hedberg, Michele Polese, Chris Dick, Tommaso Melodia

Published: 2026-11-11T05:00:00Z

Content type: article

Language: en

Sources: [arXiv: Networking](<https://devfeed.tech/sources/arxiv-networking.md>)

Topics: [real-time](<https://devfeed.tech/topics/real-time.md>), [TensorRT](<https://devfeed.tech/topics/tensorrt.md>), [network-intelligence](<https://devfeed.tech/topics/network-intelligence.md>), [Transformers](<https://devfeed.tech/topics/transformers.md>), [Nvidia](<https://devfeed.tech/topics/nvidia.md>)

Tags: [5g](<https://devfeed.tech/tags/5g.md>), [6g](<https://devfeed.tech/tags/6g.md>), [ai](<https://devfeed.tech/tags/ai.md>), [ai-applications](<https://devfeed.tech/tags/ai-applications.md>), [ai-logic](<https://devfeed.tech/tags/ai-logic.md>), [ai-native](<https://devfeed.tech/tags/ai-native.md>), [artificial-intelligence](<https://devfeed.tech/tags/artificial-intelligence.md>), [artificial-intelligence-ai](<https://devfeed.tech/tags/artificial-intelligence-ai.md>), [backends](<https://devfeed.tech/tags/backends.md>), [bandwidth](<https://devfeed.tech/tags/bandwidth.md>), [cells](<https://devfeed.tech/tags/cells.md>), [cellular](<https://devfeed.tech/tags/cellular.md>), [cs-ni](<https://devfeed.tech/tags/cs-ni.md>), [dapps](<https://devfeed.tech/tags/dapps.md>), [edge](<https://devfeed.tech/tags/edge.md>), [edge-inference](<https://devfeed.tech/tags/edge-inference.md>), [gpu-accelerated](<https://devfeed.tech/tags/gpu-accelerated.md>), [inference](<https://devfeed.tech/tags/inference.md>), [nvidia](<https://devfeed.tech/tags/nvidia.md>), [real-time](<https://devfeed.tech/tags/real-time.md>), [testbed](<https://devfeed.tech/tags/testbed.md>)

### AI overview

The paper presents an open-source framework that feeds physical-layer signals from an NVIDIA Aerial Testbed GPU-accelerated gNB to edge AI applications using the Open RAN dApp architecture. It reports about 84 μs of framework latency per slot and about 123 μs for control loops with lightweight TensorRT inference, and evaluates two GPU platforms under different sharing modes as the number of cells and dApps increases. Its cuSense demonstration uses uplink DMRS channel estimates, removes static multipath, and applies a neural network to locate a moving person without dedicated sensing hardware or changes to 5G signals. On a 3GPP-compliant 5G NR deployment, cuSense reports 73 cm mean error on runs unseen during training, 79% of predictions within 1 m, and 241 μs live runtime. The framework and dataset are released as open source.

### Source excerpt

arXiv:2512.06493v3 Announce Type: replace Abstract: The transition of cellular networks to (i) software-based systems on commodity hardware and (ii) platforms for services beyond connectivity introduces critical system-level challenges. As sensing emerges as a key feature toward 6G standardization, supporting Integrated Sensing and Communication (ISAC) with limited bandwidth and piggybacking on communication signals, while maintaining high reliability and performance, remains a fundamental challenge. In this paper, we provide two key contributions. First, we present a programmable, open-source framework for processing PHY signals through real-time, GPU-accelerated Artificial Intelligence (AI) applications on the edge Radio Access Network (RAN) infrastructure. Building on the Open RAN dApp architecture, the framework interfaces with a GPU-accelerated gNB based on NVIDIA Aerial Testbed (ATB), feeding PHY data to custom AI logic with about 84 $\mu$s of framework latency per slot and control loops with inference in about 123 $\mu$s for lightweight TensorRT models, multiple inference engines, and support for several AI backends. We evaluate the framework on two GPU platforms, with different GPU sharing modes and an increasing number of cells and dApps. Second, we demonstrate the framework capabilities through cuSense, an indoor localization dApp that consumes uplink DMRS channel estimates, removes static multipath components, and runs a neural network to infer the position of a moving person. Evaluated on a 3GPP-compliant 5G NR deployment, cuSense achieves a mean localization error of 73 cm on runs unseen during training, with 79% of predictions falling within 1 meter, and runs live in 241 $\mu$s, without dedicated sensing hardware or modifications to the 5G signals. The framework is released as open source, together with the cuSense dataset, providing a reference design for future AI-native RANs and ISAC applications.

## Robust & Scalable Website Fingerprinting via Minimal DRAM Traffic

DevFeed: [Robust & Scalable Website Fingerprinting via Minimal DRAM Traffic](<https://devfeed.tech/articles/robust-scalable-website-fingerprinting-via-minimal-dram-traffic-122802.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2611.09422>)

Author: Shubham Roy, Samar Parwez, Biswabandan Panda

Published: 2026-11-09T05:00:00Z

Content type: article

Language: en

Sources: [arXiv: Cryptography and Security](<https://devfeed.tech/sources/arxiv-cryptography-and-security.md>)

Topics: [interception](<https://devfeed.tech/topics/interception.md>), [network-intelligence](<https://devfeed.tech/topics/network-intelligence.md>), [Internet Traffic](<https://devfeed.tech/topics/internet-traffic.md>), [Website](<https://devfeed.tech/topics/website.md>)

Tags: [attack](<https://devfeed.tech/tags/attack.md>), [browsers](<https://devfeed.tech/tags/browsers.md>), [classification](<https://devfeed.tech/tags/classification.md>), [core](<https://devfeed.tech/tags/core.md>), [cs-ar](<https://devfeed.tech/tags/cs-ar.md>), [cs-cr](<https://devfeed.tech/tags/cs-cr.md>), [dram](<https://devfeed.tech/tags/dram.md>), [evaluation](<https://devfeed.tech/tags/evaluation.md>), [exploits](<https://devfeed.tech/tags/exploits.md>), [hardware](<https://devfeed.tech/tags/hardware.md>), [isolation](<https://devfeed.tech/tags/isolation.md>), [memory](<https://devfeed.tech/tags/memory.md>), [network](<https://devfeed.tech/tags/network.md>), [noise](<https://devfeed.tech/tags/noise.md>), [open](<https://devfeed.tech/tags/open.md>), [operating-systems](<https://devfeed.tech/tags/operating-systems.md>), [performance](<https://devfeed.tech/tags/performance.md>), [privacy](<https://devfeed.tech/tags/privacy.md>), [robust](<https://devfeed.tech/tags/robust.md>), [robustness](<https://devfeed.tech/tags/robustness.md>), [traffic](<https://devfeed.tech/tags/traffic.md>), [website](<https://devfeed.tech/tags/website.md>)

### AI overview

Memory-chasing fingerprints visited websites by using low-interference pointer-chasing traversals to exploit DRAM traffic interference, then counting DRAM accesses at fixed intervals to form classification traces. The abstract reports 96% accuracy on a 24-core system, 86% across multiple Core Complex Dies, 80% across multiple sockets, and 97% in an open-world setting. It contrasts this scalability with loop-counting, which falls from 93% accuracy on 8 cores to 50% on 24 cores, and sweep-counting, which falls from 84% across multiple CCDs to 45%. The authors say they evaluated robustness across hardware, browsers, operating systems, isolation mechanisms, system noise, and countermeasures; the excerpt gives no details on those conditions or results.

### Source excerpt

arXiv:2611.09422v1 Announce Type: new Abstract: Website fingerprinting attacks infer websites visited by users and threaten their privacy. These attacks are classified as on-path, which monitor network traffic, and co-located, which exploit shared system resources. Existing co-located attacks, such as sweep-counting and loop-counting, achieve over 90% accuracy but suffer significant performance degradation on many-core systems. Loop-counting accuracy drops from 93% on 8 cores to 50% on 24 cores, while sweep-counting accuracy decreases from 84% to 45% across multiple Core Complex Dies (CCDs). To address these limitations, we propose memory-chasing, a scalable website fingerprinting attack that exploits DRAM traffic interference through pointer-chasing traversals with minimal interference. By counting DRAM accesses over fixed intervals, the attack generates distinctive traces for website classification. Evaluated on the top 50 websites, memory-chasing achieves 96% accuracy on a 24-core system, 86% across multiple CCDs, 80% across multiple sockets, and 97% in an open-world setting. Extensive evaluation across hardware configurations, browsers, operating systems, isolation mechanisms, system noise, and countermeasures demonstrates its robustness and scalability.

## SEROA: Stable Encrypted-Traffic Representation and Online Adaptation for Service-Aware Networks

DevFeed: [SEROA: Stable Encrypted-Traffic Representation and Online Adaptation for Service-Aware Networks](<https://devfeed.tech/articles/seroa-stable-encrypted-traffic-representation-and-online-adaptation-for-service-aware-networks-121327.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2611.07550>)

Author: Zongxi Liu, Shuai Mao, Jie Chuai, Hongyang Du

Published: 2026-11-06T05:00:00Z

Content type: article

Language: en

Sources: [arXiv: Networking](<https://devfeed.tech/sources/arxiv-networking.md>)

Topics: [network-intelligence](<https://devfeed.tech/topics/network-intelligence.md>), [Internet Traffic](<https://devfeed.tech/topics/internet-traffic.md>), [Representation learning](<https://devfeed.tech/topics/representation-learning.md>)

Tags: [adversarial](<https://devfeed.tech/tags/adversarial.md>), [applications](<https://devfeed.tech/tags/applications.md>), [bandit](<https://devfeed.tech/tags/bandit.md>), [bandit-learning](<https://devfeed.tech/tags/bandit-learning.md>), [classification](<https://devfeed.tech/tags/classification.md>), [contrastive-learning](<https://devfeed.tech/tags/contrastive-learning.md>), [cs-ni](<https://devfeed.tech/tags/cs-ni.md>), [deployment](<https://devfeed.tech/tags/deployment.md>), [encrypted](<https://devfeed.tech/tags/encrypted.md>), [encrypted-traffic](<https://devfeed.tech/tags/encrypted-traffic.md>), [encryption](<https://devfeed.tech/tags/encryption.md>), [evaluation](<https://devfeed.tech/tags/evaluation.md>), [experiments](<https://devfeed.tech/tags/experiments.md>), [exploits](<https://devfeed.tech/tags/exploits.md>), [feedback](<https://devfeed.tech/tags/feedback.md>), [fields](<https://devfeed.tech/tags/fields.md>), [networks](<https://devfeed.tech/tags/networks.md>), [online](<https://devfeed.tech/tags/online.md>), [service](<https://devfeed.tech/tags/service.md>), [stable](<https://devfeed.tech/tags/stable.md>), [traffic](<https://devfeed.tech/tags/traffic.md>)

### AI overview

SEROA addresses temporal drift in encrypted-traffic classification for service-aware mobile networks with an offline representation-learning stage and online adaptation. Offline, it models stable protocol fields, applies protocol-aware transformations and contrastive learning to unstable fields, and uses time-adversarial learning to suppress time-specific cues. Online, it updates the classifier through contextual-bandit learning from feedback records containing QoE measurements, wireless states, and allocation decisions. The abstract reports that SEROA outperforms representative pretrained baselines under random and time-based evaluation and remains effective with QoE-driven adaptation; it gives no quantitative results or implementation details.

### Source excerpt

arXiv:2611.07550v1 Announce Type: new Abstract: Mobile networks are evolving toward service-aware operation to support heterogeneous applications with diverse quality-ofservice (QoS) and quality-of-experience (QoE) requirements. In practice, however, the widespread use of encryption makes service types difficult to observe directly, which limits the deployment of service-aware optimization in operational networks. Recent raw-byte representation learning methods have improved offline encrypted-traffic classification, but their robustness in realistic deployment remains limited due to temporal drift and the lack of explicit labels for online adaptation. This paper proposes Stable Encrypted-Traffic Representation and Online Adaptation (SEROA), a deployment-oriented framework for encryptedtraffic classification in service-aware mobile networks. In the offline stage, SEROA exploits protocol structure to learn temporally stable representations by applying masked modeling to stable fields, protocolaware transformations and contrastive learning to unstable fields, and time-adversarial learning to suppress time-specific cues. In the online stage, SEROA uses feedback records containing QoE measurements, wireless states, and allocation decisions to update the deployed classifier through contextual-bandit learning. Experiments show that SEROA consistently outperforms representative pretrained baselines under both random and time-based evaluation protocols, and remains effective under QoE-driven online adaptation. These results demonstrate the value of stable representation learning and online adaptation for practical service-aware mobile networks.

## Global Surveillance Under the Lens: Reconstructing Passive Dolev-Yao Intruders

DevFeed: [Global Surveillance Under the Lens: Reconstructing Passive Dolev-Yao Intruders](<https://devfeed.tech/articles/global-surveillance-under-the-lens-reconstructing-passive-dolev-yao-intruders-119907.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2611.07170>)

Author: Paolo Modesti

Published: 2026-11-06T05:00:00Z

Content type: article

Language: en

Sources: [arXiv: Cryptography and Security](<https://devfeed.tech/sources/arxiv-cryptography-and-security.md>), [arXiv: Software Engineering](<https://devfeed.tech/sources/arxiv-software-engineering.md>)

Topics: [Cryptography](<https://devfeed.tech/topics/cryptography.md>), [TCP/IP](<https://devfeed.tech/topics/tcp-ip.md>), [Internet Traffic](<https://devfeed.tech/topics/internet-traffic.md>), [network-intelligence](<https://devfeed.tech/topics/network-intelligence.md>), [test](<https://devfeed.tech/topics/test.md>), [Automated reasoning](<https://devfeed.tech/topics/automated-reasoning.md>)

Tags: [automated](<https://devfeed.tech/tags/automated.md>), [capture](<https://devfeed.tech/tags/capture.md>), [challenges](<https://devfeed.tech/tags/challenges.md>), [channel](<https://devfeed.tech/tags/channel.md>), [compiler](<https://devfeed.tech/tags/compiler.md>), [concurrent](<https://devfeed.tech/tags/concurrent.md>), [cryptographic](<https://devfeed.tech/tags/cryptographic.md>), [cs-cr](<https://devfeed.tech/tags/cs-cr.md>), [cs-se](<https://devfeed.tech/tags/cs-se.md>), [defensive](<https://devfeed.tech/tags/defensive.md>), [deployment](<https://devfeed.tech/tags/deployment.md>), [execution](<https://devfeed.tech/tags/execution.md>), [formal](<https://devfeed.tech/tags/formal.md>), [framework](<https://devfeed.tech/tags/framework.md>), [intruder](<https://devfeed.tech/tags/intruder.md>), [java](<https://devfeed.tech/tags/java.md>), [jdk](<https://devfeed.tech/tags/jdk.md>), [mapping](<https://devfeed.tech/tags/mapping.md>), [metadata](<https://devfeed.tech/tags/metadata.md>), [methodology](<https://devfeed.tech/tags/methodology.md>), [multiple](<https://devfeed.tech/tags/multiple.md>), [network](<https://devfeed.tech/tags/network.md>)

### AI overview

The framework reconstructs passive Dolev-Yao intruder executions from PCAP traffic and checks them against AnBx protocol specifications. Given a capture, specification and deployment metadata, it extracts Java-serialized objects from TCP streams, assigns channels to protocol steps and sessions, then runs compiler-generated semantic checks. It addresses multiple objects and references in streams, ambiguous mappings of interleaved connections to concurrent sessions, and validation of candidate reconstructions. Across 121 AnBx test-suite protocols, the standard JDK cryptographic provider recovered 97.8% of expected objects and achieved 94.2% successful reconstruction for single sessions; success required complete object recovery and a clean intruder execution. For 112 parallel-capable protocols, two- and three-session reconstructions each succeeded 89.3% of the time, with about 97.2% object recovery. Tests with role mismatches and mutated traces showed the checks detected semantic inconsistencies.

### Source excerpt

arXiv:2611.07170v1 Announce Type: new Abstract: Global surveillance systems record large volumes of network traffic for retrospective analysis, but packet captures alone reveal little about protocol-level semantics. This paper presents an automated, specification-guided methodology for reconstructing passive Dolev-Yao intruder executions from captured network traffic and validating the reconstructed executions against a formal protocol specification. The framework takes a PCAP capture, an AnBx specification, and the associated deployment metadata; it extracts Java-serialised objects from TCP streams, reconstructs protocol sessions and channel-to-step assignments, and generates a passive intruder that evaluates the Dolev-Yao derived checks produced by the AnBx compiler. Three reconstruction challenges are addressed: multiple serialised objects and object references within a TCP stream, ambiguity in mapping interleaved connections to concurrent protocol sessions, and semantic validation of candidate reconstructions. We evaluate the methodology on 121 protocols from the AnBx test suite. For single-session captures, the standard JDK cryptographic provider recovers 97.8% of expected objects and achieves 94.2% successful reconstruction, where success requires complete object recovery and a clean passive-intruder execution. For the 112 parallel-capable protocols, the toolchain achieves 89.3% successful reconstruction for both two and three concurrent sessions, while recovering approximately 97.2% of expected objects. Experiments with mismatched protocol-role mappings and deliberately mutated traces show that the generated Dolev-Yao checks respond to semantic inconsistencies. By making the reconstruction step explicit, the solution produces protocol-level observations on which assessments of protocol resilience and the design of defensive countermeasures can be built.

## PCAP-LM: Helping AI assistants read network traffic

DevFeed: [PCAP-LM: Helping AI assistants read network traffic](<https://devfeed.tech/articles/pcap-lm-helping-ai-assistants-read-network-traffic-121502.md>)

Original publisher: [Read original article](<https://blog.apnic.net/2026/11/06/pcap-lm-helping-ai-assistants-read-network-traffic/>)

Author: Xavier Marjou

Published: 2026-11-06T00:04:21Z

Content type: article

Language: en

Sources: [APNIC Blog](<https://devfeed.tech/sources/apnic-blog.md>)

Topics: [Artificial Intelligence](<https://devfeed.tech/topics/ai.md>), [network-intelligence](<https://devfeed.tech/topics/network-intelligence.md>), [Internet Traffic](<https://devfeed.tech/topics/internet-traffic.md>), [AI Chat](<https://devfeed.tech/topics/ai-chat.md>), [Network Operations](<https://devfeed.tech/topics/network-operations.md>)

Tags: [4g](<https://devfeed.tech/tags/4g.md>), [5g](<https://devfeed.tech/tags/5g.md>), [agentic](<https://devfeed.tech/tags/agentic.md>), [agentic-workflows](<https://devfeed.tech/tags/agentic-workflows.md>), [ai](<https://devfeed.tech/tags/ai.md>), [ai-assistants](<https://devfeed.tech/tags/ai-assistants.md>), [ai-model](<https://devfeed.tech/tags/ai-model.md>), [ai-services](<https://devfeed.tech/tags/ai-services.md>), [analysis](<https://devfeed.tech/tags/analysis.md>), [analyst](<https://devfeed.tech/tags/analyst.md>), [applications](<https://devfeed.tech/tags/applications.md>), [artificial-intelligence](<https://devfeed.tech/tags/artificial-intelligence.md>), [guest-post](<https://devfeed.tech/tags/guest-post.md>), [measurement](<https://devfeed.tech/tags/measurement.md>), [network](<https://devfeed.tech/tags/network.md>), [network-operations](<https://devfeed.tech/tags/network-operations.md>), [network-traffic](<https://devfeed.tech/tags/network-traffic.md>), [operations](<https://devfeed.tech/tags/operations.md>), [packet](<https://devfeed.tech/tags/packet.md>), [pcap](<https://devfeed.tech/tags/pcap.md>), [scale](<https://devfeed.tech/tags/scale.md>), [tech-matters](<https://devfeed.tech/tags/tech-matters.md>), [traffic](<https://devfeed.tech/tags/traffic.md>)

### AI overview

PCAP-LM makes packet captures small enough for an AI model to read in full by translating each conversation into compact PacketGlyphs that preserve direction, type, size, and timing while dropping payloads and other detail. It folds repeated patterns and combines frequent symbols. In tests on 30 captures from a 5G/4G measurement campaign, the rewritten files averaged about 800 times smaller than detailed Wireshark command-line text, and a model answered ten forensic questions with 99.3% accuracy. Conventional text truncated to the same length yielded 51%, partly because it exposed only the first few dozen packets. The tests covered large encrypted downloads from one server; retransmissions were missed in about a quarter of affected captures. PCAP-LM omits payload contents, sequence numbers, and exact timings, and encrypted content remains opaque. A side-index can point analysts or AI tools to original frames for follow-up. The excerpt proposes local address and hostname substitution before sharing and suggests testing other protocols, which may require new symbols or vocabulary training.

### Source excerpt

Guest Post: Packet captures are the ground truth of network operations, but analysing them is slow, requires specialist expertise, and does not scale. PCAP-LM rewrites captures into a form that AI assistants can read, helping address that mismatch.

## When Relationships Break: Interpreting Network Traffic Anomalies via Dependency Violations

DevFeed: [When Relationships Break: Interpreting Network Traffic Anomalies via Dependency Violations](<https://devfeed.tech/articles/when-relationships-break-interpreting-network-traffic-anomalies-via-dependency-violations-118675.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2608.26831>)

Author: Federica Uccello, Simin Nadjm-Tehrani

Published: 2026-11-05T05:00:00Z

Content type: article

Language: en

Sources: [arXiv: Cryptography and Security](<https://devfeed.tech/sources/arxiv-cryptography-and-security.md>)

Topics: [threat detection](<https://devfeed.tech/topics/threat-detection.md>), [outlier-detection](<https://devfeed.tech/topics/outlier-detection.md>), [network-intelligence](<https://devfeed.tech/topics/network-intelligence.md>), [netflow](<https://devfeed.tech/topics/netflow.md>), [Internet Traffic](<https://devfeed.tech/topics/internet-traffic.md>), [Incident response](<https://devfeed.tech/topics/incident-response.md>), [causality](<https://devfeed.tech/topics/causality.md>)

Tags: [attack](<https://devfeed.tech/tags/attack.md>), [cs-cr](<https://devfeed.tech/tags/cs-cr.md>), [dependency](<https://devfeed.tech/tags/dependency.md>), [detection](<https://devfeed.tech/tags/detection.md>), [features](<https://devfeed.tech/tags/features.md>), [interpretability](<https://devfeed.tech/tags/interpretability.md>), [machine-learning](<https://devfeed.tech/tags/machine-learning.md>), [modeling](<https://devfeed.tech/tags/modeling.md>), [monitoring](<https://devfeed.tech/tags/monitoring.md>), [network](<https://devfeed.tech/tags/network.md>), [network-flow](<https://devfeed.tech/tags/network-flow.md>), [patterns](<https://devfeed.tech/tags/patterns.md>), [recall](<https://devfeed.tech/tags/recall.md>), [security](<https://devfeed.tech/tags/security.md>), [security-monitoring](<https://devfeed.tech/tags/security-monitoring.md>), [shap](<https://devfeed.tech/tags/shap.md>), [temporal](<https://devfeed.tech/tags/temporal.md>), [time](<https://devfeed.tech/tags/time.md>), [traffic](<https://devfeed.tech/tags/traffic.md>)

### AI overview

XION detects network traffic anomalies by learning dependencies among flow features from benign traffic and flagging violations. After an alert, it traces which relationships broke, when they broke during an attack, and how violations evolved relative to one another. On standard IDS datasets across volumetric and stealthier attack scenarios, XION matched or exceeded Isolation Forest recall in all evaluated scenarios and used up to seven times less inference time. Its post-alert patterns were consistent with known attack behaviors and were evaluated against SHAP feature attribution for Isolation Forest alerts. These findings suggest dependency violations can add information for understanding alerts; the supplied excerpt does not detail datasets, attack-specific results, or implementation requirements.

### Source excerpt

arXiv:2608.26831v2 Announce Type: replace Abstract: Current research on security monitoring is increasingly focusing on machine-learning-based approaches, but caveats remain. In addition to huge computational overhead, one concern is the lack of insights into ``why" alerts are raised. Existing interpretability approaches rely on feature attribution methods that ignore dependencies among features or on causal modeling that requires extensive domain knowledge or computational resources. This work proposes XION, a method for modeling relationships among network-flow features based on benign traffic only. During detection, anomalies are identified through violations of expected feature dependencies. Further, XION supports post-alert analysis by identifying which feature relationships break, when they break along the attack timeline, and how dependency violations evolve relatively to other identified violations. XION is evaluated on standard IDS datasets and compared against an Isolation Forest (IF) baseline across multiple attack scenarios, including both volumetric and stealthier attacks. Results show that XION matches or exceeds IF recall in all evaluated scenarios, while requiring up to 7$\times$ less inference time. At the post-alert stage, the dependency violation analysis reveals temporal and structural patterns consistent with known attack behaviors, and is further evaluated against SHAP-based feature attribution of IF alerts. Together, these findings confirm that attacks indeed disrupt feature dependencies learned from benign traffic, and that these disruptions provide additional information for understanding an alert.

## Things that Matter -- Identifying Interactions and IoT Device Types in Encrypted Matter Traffic

DevFeed: [Things that Matter -- Identifying Interactions and IoT Device Types in Encrypted Matter Traffic](<https://devfeed.tech/articles/things-that-matter-identifying-interactions-and-iot-device-types-in-encrypted-matter-traffic-118662.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2602.01932>)

Author: Kristopher Alex Schlett, Bela Genge, Ioan Padurean, Savio Sciancalepore

Published: 2026-11-05T05:00:00Z

Content type: article

Language: en

Sources: [arXiv: Cryptography and Security](<https://devfeed.tech/sources/arxiv-cryptography-and-security.md>)

Topics: [Matter](<https://devfeed.tech/topics/matter.md>), [Internet of things](<https://devfeed.tech/topics/iot.md>), [network-intelligence](<https://devfeed.tech/topics/network-intelligence.md>), [Cryptography](<https://devfeed.tech/topics/cryptography.md>), [interception](<https://devfeed.tech/topics/interception.md>), [Internet Traffic](<https://devfeed.tech/topics/internet-traffic.md>)

Tags: [cs-cr](<https://devfeed.tech/tags/cs-cr.md>), [devices](<https://devfeed.tech/tags/devices.md>), [encrypted-traffic](<https://devfeed.tech/tags/encrypted-traffic.md>), [iot](<https://devfeed.tech/tags/iot.md>), [lighting](<https://devfeed.tech/tags/lighting.md>), [lock](<https://devfeed.tech/tags/lock.md>), [matter](<https://devfeed.tech/tags/matter.md>), [metadata](<https://devfeed.tech/tags/metadata.md>), [plug](<https://devfeed.tech/tags/plug.md>), [privacy](<https://devfeed.tech/tags/privacy.md>), [risk](<https://devfeed.tech/tags/risk.md>), [security](<https://devfeed.tech/tags/security.md>), [sequences](<https://devfeed.tech/tags/sequences.md>)

### AI overview

A passive observer can infer Matter activity despite encrypted communications: the study reports over 95% accuracy in identifying interactions and over 88% accuracy in inferring device types, including lights, locks, and plugs. It builds repeatable fingerprints from packet lengths, directions, and sequences using data from real-world and simulated deployments; interaction identification remained above 95% under packet loss and delay. The results indicate a privacy risk across Matter deployments, and the Connectivity Standards Alliance acknowledged the findings and said mitigation may be considered in future Matter versions.

### Source excerpt

arXiv:2602.01932v3 Announce Type: replace Abstract: Matter is the latest interoperability standard for the consumer Internet of Things (IoT) market, developed by the Connectivity Standards Alliance (CSA). Matter's design pays particular attention to security and privacy: it provides validated secure session establishment protocols, and it uses robust security algorithms to encrypt communications between IoT devices and Matter controllers. However, little is known about the extent to which information can still be inferred from encrypted traffic. This paper investigates whether a passive adversary can identify Matter interactions and device types by analyzing encrypted traffic metadata. Using datasets collected from both real-world and simulated Matter deployments, we identify repeatable fingerprints based on packet lengths, directions, and packet sequences. Our results show that Matter interactions can be identified with over 95% accuracy even under packet loss and delay conditions. Furthermore, device types such as lighting, lock, and plug appliances can be inferred with accuracy exceeding 88%. These findings reveal a significant privacy risk affecting Matter-enabled deployments regardless of vendor. The CSA acknowledged our findings and indicated that mitigation mechanisms may be considered in future Matter versions.

## Explainable Anomaly Detection in 5G Networks through Continuous Time Graphs

DevFeed: [Explainable Anomaly Detection in 5G Networks through Continuous Time Graphs](<https://devfeed.tech/articles/explainable-anomaly-detection-in-5g-networks-through-continuous-time-graphs-118705.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2611.06225>)

Author: Thomas Hoger (LAAS, LAAS-SARA, LAAS-RISC), Philippe Owezarski (LAAS, LAAS-SARA, LAAS-RISC)

Published: 2026-11-05T05:00:00Z

Content type: article

Language: en

Sources: [arXiv: Cryptography and Security](<https://devfeed.tech/sources/arxiv-cryptography-and-security.md>), [arXiv: Networking](<https://devfeed.tech/sources/arxiv-networking.md>)

Topics: [threat detection](<https://devfeed.tech/topics/threat-detection.md>), [network security](<https://devfeed.tech/topics/network-security.md>), [Graph Mining](<https://devfeed.tech/topics/graph-mining.md>), [network-intelligence](<https://devfeed.tech/topics/network-intelligence.md>)

Tags: [5g](<https://devfeed.tech/tags/5g.md>), [anomaly-detection](<https://devfeed.tech/tags/anomaly-detection.md>), [attacks](<https://devfeed.tech/tags/attacks.md>), [cs-cr](<https://devfeed.tech/tags/cs-cr.md>), [cs-ne](<https://devfeed.tech/tags/cs-ne.md>), [cs-ni](<https://devfeed.tech/tags/cs-ni.md>), [detection](<https://devfeed.tech/tags/detection.md>), [distributed](<https://devfeed.tech/tags/distributed.md>), [graphs](<https://devfeed.tech/tags/graphs.md>), [networks](<https://devfeed.tech/tags/networks.md>)

### AI overview

The paper models 5G network exchanges as a dynamic graph and applies graph-based learning to detect anomalies. The representation is intended to capture packet-level semantic content alongside sequential and topological context, which the authors say can help identify sophisticated, multi-stage attacks. They evaluate the approach using popular dynamic graph models; the supplied excerpt reports potential for 5G anomaly detection but gives no model names or performance results.

### Source excerpt

arXiv:2611.06225v1 Announce Type: new Abstract: 5G networks increasingly resemble distributed systems in their design. It then becomes crucial to move beyond traditional network-based detection methods and to explore alternative approaches. In this work, we propose a formalization of 5G network exchanges as a dynamic graph, on which we apply techniques from the field of graph-based learning to detect anomalies effectively. This representation allows for fine-grained analysis of packet-level semantic content, while simultaneously capturing sequential and topological contexts, both of which are essential for identifying sophisticated and multi-stage attacks. To assess the effectiveness of this formalism, we evaluate it with the most popular dynamic graph models, demonstrating its potential for 5G anomaly detection.

## Survival-Time Risk Beyond Low-Order Channel Models: Identifiability and Scheduling

DevFeed: [Survival-Time Risk Beyond Low-Order Channel Models: Identifiability and Scheduling](<https://devfeed.tech/articles/survival-time-risk-beyond-low-order-channel-models-identifiability-and-scheduling-116517.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2611.03871>)

Author: Ruifeng Zheng, Zhihan Xu, Pit Hofmann, Frank H. P. Fitzek

Published: 2026-11-04T05:00:00Z

Content type: article

Language: en

Sources: [arXiv: Networking](<https://devfeed.tech/sources/arxiv-networking.md>)

Topics: [network-intelligence](<https://devfeed.tech/topics/network-intelligence.md>), [scheduling](<https://devfeed.tech/topics/scheduling.md>)

Tags: [channel](<https://devfeed.tech/tags/channel.md>), [cs-it](<https://devfeed.tech/tags/cs-it.md>), [cs-ni](<https://devfeed.tech/tags/cs-ni.md>), [markov](<https://devfeed.tech/tags/markov.md>), [math-it](<https://devfeed.tech/tags/math-it.md>), [memory](<https://devfeed.tech/tags/memory.md>), [models](<https://devfeed.tech/tags/models.md>), [order](<https://devfeed.tech/tags/order.md>), [probability](<https://devfeed.tech/tags/probability.md>), [relapse](<https://devfeed.tech/tags/relapse.md>), [retry](<https://devfeed.tech/tags/retry.md>), [reverse](<https://devfeed.tech/tags/reverse.md>), [risk](<https://devfeed.tech/tags/risk.md>), [scheduling](<https://devfeed.tech/tags/scheduling.md>), [spacing](<https://devfeed.tech/tags/spacing.md>), [statistics](<https://devfeed.tech/tags/statistics.md>), [survival](<https://devfeed.tech/tags/survival.md>), [time](<https://devfeed.tech/tags/time.md>), [transmission](<https://devfeed.tech/tags/transmission.md>), [update](<https://devfeed.tech/tags/update.md>)

### AI overview

Channels with the same low-order packet-loss statistics can have substantially different risks of receiving no successful update for a prescribed run of slots. The paper characterizes compatible Markov channels one order above a low-order fit and analyzes them under acknowledgment-based scheduling with transmission budgets. With dense attempts, loss persistence beyond the modeled memory largely governs risk, but the most persistent channel need not be worst. With limited budgets, retry spacing makes loss-relapse probabilities important and can reverse the risk ordering. Numerical and measurement results indicate that low-order models can substantially underestimate risk, so assessment depends on scheduling-related channel statistics.

### Source excerpt

arXiv:2611.03871v1 Announce Type: new Abstract: This paper investigates whether low-order packet-loss statistics suffice to assess survival-time risk under bursty packet losses, i.e., the probability that no update succeeds for a prescribed number of consecutive slots. Conventional models, such as the Gilbert model, characterize packet losses using the loss rate and mean burst length, but channels sharing these statistics may exhibit substantially different failure risks. We characterize the Markov channels one order above a given low-order fit that share its statistics and analyze their risks under acknowledgment-based scheduling with transmission budgets. We show that, with dense transmission attempts, risk is primarily governed by loss persistence beyond the modeled memory, and the worst-case channel need not be the most persistent one. Under limited budgets, retry spacing makes loss relapse probabilities important and can reverse the risk ordering of compatible channels. Numerical and measurement results show that low-order models can substantially underestimate the risk, highlighting the need for scheduling-dependent channel statistics.

## Which Deep Learner? A Systematic Evaluation of Advanced Deep Forecasting Models Accuracy and Efficiency for Network Traffic Prediction

DevFeed: [Which Deep Learner? A Systematic Evaluation of Advanced Deep Forecasting Models Accuracy and Efficiency for Network Traffic Prediction](<https://devfeed.tech/articles/which-deep-learner-a-systematic-evaluation-of-advanced-deep-forecasting-models-accuracy-and-efficiency-for-network-traffic-prediction-116513.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2601.02694>)

Author: Eilaf M. A. Babai, Aalaa M. A. Babai, Koji Okamura

Published: 2026-11-04T05:00:00Z

Content type: article

Language: en

Sources: [arXiv: Machine Learning](<https://devfeed.tech/sources/arxiv-machine-learning.md>), [arXiv: Networking](<https://devfeed.tech/sources/arxiv-networking.md>)

Topics: [network-intelligence](<https://devfeed.tech/topics/network-intelligence.md>), [Machine learning](<https://devfeed.tech/topics/machine-learning.md>), [Predictive AI](<https://devfeed.tech/topics/predictive-ai.md>), [Time Series](<https://devfeed.tech/topics/time-series.md>), [Data analysis](<https://devfeed.tech/topics/data-analysis.md>)

Tags: [cs-lg](<https://devfeed.tech/tags/cs-lg.md>), [cs-ni](<https://devfeed.tech/tags/cs-ni.md>), [deep-learning](<https://devfeed.tech/tags/deep-learning.md>), [deployment](<https://devfeed.tech/tags/deployment.md>), [efficiency](<https://devfeed.tech/tags/efficiency.md>), [energy](<https://devfeed.tech/tags/energy.md>), [evaluation](<https://devfeed.tech/tags/evaluation.md>), [external](<https://devfeed.tech/tags/external.md>), [forecasting](<https://devfeed.tech/tags/forecasting.md>), [memory](<https://devfeed.tech/tags/memory.md>), [models](<https://devfeed.tech/tags/models.md>), [network-management](<https://devfeed.tech/tags/network-management.md>), [network-traffic](<https://devfeed.tech/tags/network-traffic.md>), [performance](<https://devfeed.tech/tags/performance.md>), [prediction](<https://devfeed.tech/tags/prediction.md>), [robustness](<https://devfeed.tech/tags/robustness.md>), [series](<https://devfeed.tech/tags/series.md>), [time-series](<https://devfeed.tech/tags/time-series.md>), [time-series-forecasting](<https://devfeed.tech/tags/time-series-forecasting.md>)

### AI overview

The study evaluates 15 forecasting methods--12 deep models and 3 statistical baselines--on four real network traffic datasets across multiple timescales and forecast horizons. It examines accuracy alongside robustness to anomalies, data gaps and external factors, data efficiency, and time, memory, and energy use. The abstract reports that the results identify performance regimes, efficiency thresholds, and architectures that balance these considerations, but gives no model-specific rankings or findings.

### Source excerpt

arXiv:2601.02694v2 Announce Type: replace Abstract: Network traffic prediction is an essential tool for automating modern network management. It is a difficult time series forecasting (TSF) problem that has been addressed by Deep Learning (DL) models due to their ability to capture complex patterns. Recent years have witnessed advances in deep forecasting models, from sophisticated transformer architectures to simple linear models, improving performance across diverse prediction tasks. However, given the diversity of these models, the variability of network traffic across environments and timescales, and the inherent challenges of network traffic prediction, it is essential to identify effective deployment choices and modeling directions. Henceforth, this study systematically identifies 12 deep forecasting models, comprising 8 advanced TSF models and 4 traditional deep learning architectures -- each with unique advantages for network traffic prediction -- against 3 statistical baselines, for a total of 15 forecasting methods and evaluate them on four real traffic datasets, across multiple time scales and horizons, assessing performance, robustness to anomalies, data gaps, external factors, data efficiency, and resource efficiency in terms of time, memory, and energy. Our results provide insights into model behaviour and guidelines for network management, highlighting performance regimes, efficiency thresholds, and promising architectures that balance accuracy, efficiency, and robustness to traffic challenges.

## MLEP in KICS for Networks 5.0: how machine learning helps track network behavior of OT devices

DevFeed: [MLEP in KICS for Networks 5.0: how machine learning helps track network behavior of OT devices](<https://devfeed.tech/articles/mlep-in-kics-for-networks-5-0-how-machine-learning-helps-track-network-behavior-of-ot-devices-116001.md>)

Original publisher: [Read original article](<https://securelist.com/tr/machine-learning-event-processor-in-kics-for-networks/121794/>)

Author: Roman Ezhov, Andrey Lavrentyev

Published: 2026-11-03T10:00:07Z

Content type: article

Language: en

Sources: [Securelist](<https://devfeed.tech/sources/securelist.md>)

Topics: [Machine learning](<https://devfeed.tech/topics/machine-learning.md>), [Artificial Intelligence](<https://devfeed.tech/topics/ai.md>), [Neural Network](<https://devfeed.tech/topics/neural-network.md>), [outlier-detection](<https://devfeed.tech/topics/outlier-detection.md>), [network-intelligence](<https://devfeed.tech/topics/network-intelligence.md>)

Tags: [ai](<https://devfeed.tech/tags/ai.md>), [alert](<https://devfeed.tech/tags/alert.md>), [anomaly-detection](<https://devfeed.tech/tags/anomaly-detection.md>), [asset-discovery](<https://devfeed.tech/tags/asset-discovery.md>), [asset-management](<https://devfeed.tech/tags/asset-management.md>), [asset-visibility](<https://devfeed.tech/tags/asset-visibility.md>), [assets](<https://devfeed.tech/tags/assets.md>), [attack](<https://devfeed.tech/tags/attack.md>), [behavior](<https://devfeed.tech/tags/behavior.md>), [communication](<https://devfeed.tech/tags/communication.md>), [cybersecurity](<https://devfeed.tech/tags/cybersecurity.md>), [devices](<https://devfeed.tech/tags/devices.md>), [event](<https://devfeed.tech/tags/event.md>), [full](<https://devfeed.tech/tags/full.md>), [industrial](<https://devfeed.tech/tags/industrial.md>), [industrial-control-systems](<https://devfeed.tech/tags/industrial-control-systems.md>), [industrial-threats](<https://devfeed.tech/tags/industrial-threats.md>), [inventory](<https://devfeed.tech/tags/inventory.md>), [kaspersky](<https://devfeed.tech/tags/kaspersky.md>), [large](<https://devfeed.tech/tags/large.md>), [machine-learning](<https://devfeed.tech/tags/machine-learning.md>), [medium](<https://devfeed.tech/tags/medium.md>), [networks](<https://devfeed.tech/tags/networks.md>), [neural-networks](<https://devfeed.tech/tags/neural-networks.md>), [operational-technology](<https://devfeed.tech/tags/operational-technology.md>), [ot](<https://devfeed.tech/tags/ot.md>), [security-technology](<https://devfeed.tech/tags/security-technology.md>), [threat-hunting](<https://devfeed.tech/tags/threat-hunting.md>), [thumbnail](<https://devfeed.tech/tags/thumbnail.md>)

### AI overview

Kaspersky Industrial CyberSecurity for Networks 5.0 uses MLEP to refine OT device categories from recurring network behavior and flag deviations from each device's usual profile. It represents traffic as timestamped 15-second session fragments containing endpoint, direction, protocol, and packet count and size, then processes event streams with a spiking neural semantic network. For categorization, it groups devices by shared behavior, uses known Asset Discovery categories to suggest categories for uncategorized devices, and checks suggestions against available metadata and rules. Preconfigured monitors alert on behavioral deviations such as new connections, protocol changes, transfer volume, or operating hours; analysts can correlate these alerts with device roles and other events. MLEP complements passive Asset Discovery and optional active polling, whose available details depend on protocols, observed traffic, and polling suitability. Its aggregated events cannot reconstruct packet order or command content, and results depend on traffic visibility, especially for rarely communicating devices.

### Source excerpt

We explain how Kaspersky Industrial CyberSecurity for Networks uses Machine Learning Event Processor (MLEP) technology to help take inventory of assets and detect anomalies in industrial networks.

## ICS-Sniper: A Targeted Blackhole Attack on Encrypted ICS Traffic

DevFeed: [ICS-Sniper: A Targeted Blackhole Attack on Encrypted ICS Traffic](<https://devfeed.tech/articles/ics-sniper-a-targeted-blackhole-attack-on-encrypted-ics-traffic-113965.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2312.06140>)

Author: Gargi Mitra, Chanyuan Liu, Pritam Dash, Yingao Elaine Yao, Alain Zhiyanov, Aastha Mehta, Karthik Pattabiraman

Published: 2026-11-03T05:00:00Z

Content type: article

Language: en

Sources: [arXiv: Cryptography and Security](<https://devfeed.tech/sources/arxiv-cryptography-and-security.md>)

Topics: [Security](<https://devfeed.tech/topics/security.md>), [network-intelligence](<https://devfeed.tech/topics/network-intelligence.md>), [Virtual Private Network](<https://devfeed.tech/topics/vpn.md>), [interception](<https://devfeed.tech/topics/interception.md>), [End-to-End Encryption](<https://devfeed.tech/topics/end-to-end-encryption.md>), [payload](<https://devfeed.tech/topics/payload.md>)

Tags: [abstract](<https://devfeed.tech/tags/abstract.md>), [acquisition](<https://devfeed.tech/tags/acquisition.md>), [adversary](<https://devfeed.tech/tags/adversary.md>), [analysis](<https://devfeed.tech/tags/analysis.md>), [arxiv](<https://devfeed.tech/tags/arxiv.md>), [attack](<https://devfeed.tech/tags/attack.md>), [attacks](<https://devfeed.tech/tags/attacks.md>), [automation](<https://devfeed.tech/tags/automation.md>), [blackhole](<https://devfeed.tech/tags/blackhole.md>), [can](<https://devfeed.tech/tags/can.md>), [cloud](<https://devfeed.tech/tags/cloud.md>), [commands](<https://devfeed.tech/tags/commands.md>), [communications](<https://devfeed.tech/tags/communications.md>), [contain](<https://devfeed.tech/tags/contain.md>), [control](<https://devfeed.tech/tags/control.md>), [costs](<https://devfeed.tech/tags/costs.md>), [critical](<https://devfeed.tech/tags/critical.md>), [cs-cr](<https://devfeed.tech/tags/cs-cr.md>), [encrypted](<https://devfeed.tech/tags/encrypted.md>), [ics](<https://devfeed.tech/tags/ics.md>), [traffic](<https://devfeed.tech/tags/traffic.md>)

### AI overview

ICS-Sniper shows how an on-path adversary can disrupt cloud-connected ICS despite VPN encryption and without entering the ICS perimeter or knowing its control logic. It analyzes packet sizes, directions, and timing to identify "critical superperiods," when traffic is likely to carry important commands or data, then drops a small number of payload-carrying packets during those windows in a later operational cycle. The authors report three attacks across two Secure Water Treatment (SWaT) testbeds that could threaten operational safety while evading state-of-the-art ICS attack detectors. These findings are limited to the described testbeds and demonstrations.

### Source excerpt

arXiv:2312.06140v5 Announce Type: replace Abstract: Modern industrial control systems (ICS) increasingly host their Supervisory Control and Data Acquisition (SCADA) services in the cloud to reduce the costs of large-scale automation. To protect site-SCADA communications, ICS operators commonly use VPN tunneling and standard security practices. We show that, despite these security measures, an on-path Internet adversary can disrupt ICS operations without infiltrating the ICS perimeter, breaking encryption, or knowledge of the control logic. We present ICS-Sniper, a targeted blackhole attack that analyzes the VPN traffic metadata (sizes, direction, timing of packets) to identify narrow time windows, called critical superperiods, during which the site-SCADA traffic would likely contain highly critical commands or data. Post-analysis, in a subsequent operational cycle, ICS-Sniper drops a small set of payload-carrying packets in the critical superperiods to disrupt the ICS's operations. We demonstrate three attacks on two realistic modern Secure Water Treatment (SWaT) plant testbeds that can potentially violate the operational safety of the ICS while evading state-of-the-art ICS attack detectors.

## TrafficInstruct: A Multi-Task Network Traffic Dataset and Benchmark for Instruction Tuning

DevFeed: [TrafficInstruct: A Multi-Task Network Traffic Dataset and Benchmark for Instruction Tuning](<https://devfeed.tech/articles/trafficinstruct-a-multi-task-network-traffic-dataset-and-benchmark-for-instruction-tuning-114094.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2611.03319>)

Author: Zhengdao Li, Hui Wen, Zixuan Wan, Weifeng Zhang, Yan Wu, Benlong Wu, Nenghai Yu, Weiming Zhang, Kejiang Chen

Published: 2026-11-03T05:00:00Z

Content type: article

Language: en

Sources: [arXiv: Cryptography and Security](<https://devfeed.tech/sources/arxiv-cryptography-and-security.md>)

Topics: [network-intelligence](<https://devfeed.tech/topics/network-intelligence.md>), [threat detection](<https://devfeed.tech/topics/threat-detection.md>), [Fine-tuning](<https://devfeed.tech/topics/fine-tuning.md>), [lora](<https://devfeed.tech/topics/lora.md>), [Benchmark](<https://devfeed.tech/topics/benchmark.md>)

Tags: [advanced-persistent-threat](<https://devfeed.tech/tags/advanced-persistent-threat.md>), [benchmark](<https://devfeed.tech/tags/benchmark.md>), [botnet](<https://devfeed.tech/tags/botnet.md>), [classification](<https://devfeed.tech/tags/classification.md>), [concept-drift](<https://devfeed.tech/tags/concept-drift.md>), [cs-ai](<https://devfeed.tech/tags/cs-ai.md>), [cs-cr](<https://devfeed.tech/tags/cs-cr.md>), [dataset](<https://devfeed.tech/tags/dataset.md>), [dns-over-https](<https://devfeed.tech/tags/dns-over-https.md>), [fine-tuning](<https://devfeed.tech/tags/fine-tuning.md>), [lora](<https://devfeed.tech/tags/lora.md>), [malicious](<https://devfeed.tech/tags/malicious.md>), [malware](<https://devfeed.tech/tags/malware.md>), [network-traffic](<https://devfeed.tech/tags/network-traffic.md>), [network-traffic-analysis](<https://devfeed.tech/tags/network-traffic-analysis.md>), [raw](<https://devfeed.tech/tags/raw.md>), [structured](<https://devfeed.tech/tags/structured.md>), [supervised-fine-tuning](<https://devfeed.tech/tags/supervised-fine-tuning.md>), [tor](<https://devfeed.tech/tags/tor.md>), [vpn](<https://devfeed.tech/tags/vpn.md>), [web](<https://devfeed.tech/tags/web.md>), [website](<https://devfeed.tech/tags/website.md>)

### AI overview

TrafficInstruct provides 55,000 structured training records across ten network traffic analysis tasks, alongside raw packet captures. Traffic was collected through controlled software executions, attack experiments, and application activities; labels are backed by task-specific evidence such as process or socket attribution and activity records. The dataset is filtered and transformed into instructions, traffic observations, and structured answers. Samples are partitioned by complete collection sessions, and the benchmark specifies allowed inputs, label mappings, fixed splits, and task-wise scoring. Evaluations compare majority-class, random-forest, ET-BERT, an unadapted 27B language model, and task-specific LoRA adapters on a shared frozen base. The abstract reports that task-specific adaptation improves classification performance; it does not provide metric values here.

### Source excerpt

arXiv:2611.03319v1 Announce Type: new Abstract: Instruction tuning for multi-task network traffic analysis requires data with clearly defined sources, supervision targets, and evaluation conditions. We introduce TrafficInstruct, a newly collected dataset comprising raw packet captures and 55,000 structured supervised fine-tuning records across ten tasks: malware traffic detection, botnet detection, malicious DNS-over-HTTPS detection, Web attack detection, advanced persistent threat detection, encrypted VPN detection, Tor behavior detection, encrypted application/service classification, website fingerprinting, and application classification under concept drift. An automated workflow captures traffic from controlled software executions, attack experiments, and application activities. Labels are supported by task-specific collection evidence, including process or socket attribution, activity records, and execution outcomes. Filtering, sanitization, and task-specific transformations produce instructions, traffic observations, and structured answers. We verify sample sources and partition the dataset by complete collection sessions, keeping all samples from the same session within one partition. The benchmark defines allowed inputs, deterministic label mappings, fixed training/validation/test partitions, and task-wise scoring rules. We evaluate majority-class, random-forest, and ET-BERT baselines, an unadapted 27B language model, and independently trained task-specific LoRA adapters on a shared frozen base. Adapters are selected using validation data and evaluated on the corresponding test sets. Task-specific adaptation improves classification performance, supporting instruction tuning across diverse traffic-analysis tasks.

## QueueTrack: Framework for Detecting the Onset of Queueing in Mobile Networks

DevFeed: [QueueTrack: Framework for Detecting the Onset of Queueing in Mobile Networks](<https://devfeed.tech/articles/queuetrack-framework-for-detecting-the-onset-of-queueing-in-mobile-networks-113870.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2611.02038>)

Author: Hamid Hassani, Anudeep Karnam, Kishor Joshi, Sonia M. Heemstra, Ignas G. Niemegeers, George Exarchakos

Published: 2026-11-03T05:00:00Z

Content type: article

Language: en

Sources: [arXiv: Networking](<https://devfeed.tech/sources/arxiv-networking.md>)

Topics: [Internet Traffic](<https://devfeed.tech/topics/internet-traffic.md>), [Machine learning](<https://devfeed.tech/topics/machine-learning.md>), [queueing theory](<https://devfeed.tech/topics/queueing-theory.md>), [network-intelligence](<https://devfeed.tech/topics/network-intelligence.md>)

Tags: [5g](<https://devfeed.tech/tags/5g.md>), [classification](<https://devfeed.tech/tags/classification.md>), [congestion-control](<https://devfeed.tech/tags/congestion-control.md>), [cs-ni](<https://devfeed.tech/tags/cs-ni.md>), [detection](<https://devfeed.tech/tags/detection.md>), [framework](<https://devfeed.tech/tags/framework.md>), [logistic-regression](<https://devfeed.tech/tags/logistic-regression.md>), [machine-learning](<https://devfeed.tech/tags/machine-learning.md>), [mobile-networks](<https://devfeed.tech/tags/mobile-networks.md>), [monitoring](<https://devfeed.tech/tags/monitoring.md>), [network-telemetry](<https://devfeed.tech/tags/network-telemetry.md>), [tcp](<https://devfeed.tech/tags/tcp.md>), [udp](<https://devfeed.tech/tags/udp.md>)

### AI overview

QueueTrack detects the onset of queueing at 5G base-station bottlenecks using packet-level timing statistics observable at end hosts. Its queue buildup indicator (QBI) requires neither tight clock synchronization nor changes to network infrastructure. The authors evaluate its full machine-learning pipeline on an OpenAirInterface 5G testbed across scenarios with competing TCP traffic; low-complexity models such as logistic regression detect queueing, with an F1 score of about 80% in a challenging mixed TCP/UDP scenario. The abstract presents the framework as a potential basis for low-delay congestion control and mobile-network monitoring.

### Source excerpt

arXiv:2611.02038v1 Announce Type: new Abstract: In mobile networks, timely and accurate detection of the onset of queueing at path bottlenecks remains challenging for end-to-end congestion control algorithms (CCAs), where end hosts have limited visibility into queue status. Traditional congestion signals, such as packet loss, round-trip time (RTT), and one-way delay (OWD), have well-known limitations for early queue buildup detection. This paper presents QueueTrack, a practical end-host framework to detect the onset of queueing in mobile networks. QueueTrack computes a queue buildup indicator, denoted QBI, from endpoint-observable packet-level timing statistics, without requiring tight synchronization between end hosts or modifications to the existing network infrastructure. We develop and assess the complete QueueTrack machine learning (ML) pipeline, including data collection, labeling, model training, and testing, on a 5G testbed built on the OpenAirInterface platform, under a range of scenarios with competing TCP traffic. The experimental results show that QueueTrack can reliably detect the onset of queueing at the 5G base station using low-complexity classification models such as logistic regression. For example, QueueTrack achieves an F1 score of approximately 80% in a challenging mixed TCP/UDP scenario. These findings confirm that QueueTrack can serve as a practical framework for developing CCAs that target low delay, as well as for network telemetry and network monitoring in mobile networks.

## DeepWeave: Weaving Neural and Symbolic Evidence for ML-Based Network Intrusion Detection System

DevFeed: [DeepWeave: Weaving Neural and Symbolic Evidence for ML-Based Network Intrusion Detection System](<https://devfeed.tech/articles/deepweave-weaving-neural-and-symbolic-evidence-for-ml-based-network-intrusion-detection-system-111224.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2610.41541>)

Author: Raoul Saad, Enkeleda Bardhi, Andrea Agiollo, Fernando Kuipers

Published: 2026-11-02T05:00:00Z

Content type: article

Language: en

Sources: [arXiv: Cryptography and Security](<https://devfeed.tech/sources/arxiv-cryptography-and-security.md>), [arXiv: Machine Learning](<https://devfeed.tech/sources/arxiv-machine-learning.md>), [arXiv: Networking](<https://devfeed.tech/sources/arxiv-networking.md>)

Topics: [Detection engineering](<https://devfeed.tech/topics/detection-engineering.md>), [Machine learning](<https://devfeed.tech/topics/machine-learning.md>), [Neural Network](<https://devfeed.tech/topics/neural-network.md>), [Artificial Intelligence](<https://devfeed.tech/topics/ai.md>), [network-intelligence](<https://devfeed.tech/topics/network-intelligence.md>)

Tags: [alerts](<https://devfeed.tech/tags/alerts.md>), [cs-cr](<https://devfeed.tech/tags/cs-cr.md>), [cs-lg](<https://devfeed.tech/tags/cs-lg.md>), [cs-ni](<https://devfeed.tech/tags/cs-ni.md>), [false-positive-rate](<https://devfeed.tech/tags/false-positive-rate.md>), [flow](<https://devfeed.tech/tags/flow.md>), [hybrid](<https://devfeed.tech/tags/hybrid.md>), [inference](<https://devfeed.tech/tags/inference.md>), [intrusion-detection](<https://devfeed.tech/tags/intrusion-detection.md>), [logic](<https://devfeed.tech/tags/logic.md>), [ml](<https://devfeed.tech/tags/ml.md>), [network](<https://devfeed.tech/tags/network.md>), [neural](<https://devfeed.tech/tags/neural.md>), [packet](<https://devfeed.tech/tags/packet.md>), [probabilistic](<https://devfeed.tech/tags/probabilistic.md>), [proofs](<https://devfeed.tech/tags/proofs.md>), [robustness](<https://devfeed.tech/tags/robustness.md>), [rules](<https://devfeed.tech/tags/rules.md>)

### AI overview

DeepWeave combines probabilistic neural predictions with automatically constructed and validated interpretable rules, reconciling their evidence through explicit policies across packet, flow, and hybrid views. Its proofs explain alerts supported by rules. In the reported evaluation across multiple datasets, with 50 labeled packet observations per attack class, it raised macro-F1 from 0.632 to 0.924 and lowered attack false-positive rate from 0.075 to 0.008 versus a matched neural baseline. The abstract also reports more stable proofs than post-hoc neural attributions and practical probabilistic reasoning through batched inference. These are abstract-level claims; details of datasets and evaluation are not provided in the supplied text.

### Source excerpt

arXiv:2610.41541v1 Announce Type: new Abstract: Machine learning-based network intrusion detection systems can recognize complex traffic patterns, but often require substantial labelled attack data, degrade under deployment shift, and provide limited evidence for analysts. We present DeepWeave, a probabilistic neuro-symbolic Network Intrusion Detection System (NIDS) that makes logic an integral part of detection. DeepWeave automatically constructs and validates interpretable rules, represents neural predictions as probabilistic predicates, and reconciles both evidence sources through explicit coupling policies over packet-, flow-, and hybrid-views. The resulting proofs provide intrinsic explanations for rule-supported alerts. Across multiple datasets, DeepWeave improves low-data detection, false-positive control, transferability, and robustness. With only 50 labelled packet observations per attack class, it improves macro-F1 from 0.632 to 0.924 and reduces the attack false-positive rate from 0.075 to 0.008 over the matched neural baseline. Its proofs are also more stable than post-hoc neural attributions, while batched inference keeps probabilistic reasoning practical. These results demonstrate label-efficient, adaptable, and inspectable neuro-symbolic intrusion detection.

[Next page](<https://devfeed.tech/topics/network-intelligence.md?cursor=WyIyMDI2LTExLTAyVDA1OjAwOjAwKzAwOjAwIiwgIjU3YWIzOWVlLWZmOTMtNGQ5MS1iYTI3LWQ4MmJhZDI2YmIyZSJd>)