ResLearn-XR is a residual learning framework designed to predict extended reality (XR) network traffic and estimate Quality-of-Experience (QoE) risk. It uses a two‑stage temporal learning structure: a base sequence prediction model followed by task‑specific residual components that operate in value space for traffic forecasting and in logit space for probabilistic QoE risk estimation. The framework introduces a Data Descriptor Algorithm (DDA) to convert packet‑level observables into frame‑timing‑aware descriptors and is evaluated on a newly constructed XR Traffic‑QoE dataset, achieving significant reductions in SMAPE for both traffic prediction and QoE‑risk estimation compared to single‑stage baselines.
By Yoga Suhas Kuruba Manjunath, Jie Gao, Lian Zhao
arXiv:2511. 08851v5 Announce Type: replace-cross Abstract: This paper presents a measurement-driven study of early warning for reliability breakdown events in 5G non-standalone (NSA) railway networks.
By Po-Heng Chou, Da-Chih Lin, Hung-Yu Wei, Walid Saad, Yu Tsao
The paper proposes a human-centered framework for validating the semantic soundness of machine learning models used in network traffic classification. It extends existing knowledge-generation methods by integrating data, models, explainability tools, visualizations, and expert reasoning to iteratively explore, verify, and refine model behavior and preprocessing steps. The framework is built on literature findings, benchmark analyses, XAI experience, and expert feedback, offering practical guidance for ensuring models learn trustworthy, semantically meaningful patterns rather than spurious correlations.
By Igor Cherepanov, David Sessler, Alex Ulmer, Thorsten May, J\"orn Kohlhammer
Machine learning (ML) has become the dominant approach for network traffic classification, achieving very high predictive performance. However, a model is only valuable if it learns semantically meani...
arXiv:2608. 00402v1 Announce Type: new Abstract: Estimated Time of Arrival (ETA) prediction is a core component of intelligent transportation systems.
By Yanming Lyu, Yue Cheng, Lingkun Li, Ruipeng Gao, Xinyue Liu, Hui Gao, Qiang Ni
arXiv:2607. 06979v1 Announce Type: new Abstract: Federated Learning (FL) enables training shared models on private, on-device data, but production deployments remain constrained to slow, multi-day refresh cycles due to the complexity of coordinating massive client populations.
By Dhruv Garg, Neha Lakhani, Debopam Sanyal, Myungjin Lee, Alexey Tumanov, Ada Gavrilovska
Virtual sensing, the estimation of hard-to-measure quantities from available sensor measurements, is a critical enabler for control and monitoring in cyber-physical systems. However, when sensors fail...
arXiv:2609.09442v2 Announce Type: replace-cross
Abstract: In network traffic, legitimate behaviours and attack techniques evolve jointly - the phenomenon known as 'concept drift' [1]. Every detector...
By Julien Michel, Abdul Qadir Khan, Majed Jaber, Pierre Parrend
arXiv:2605. 01989v2 Announce Type: replace Abstract: Distributed machine learning (ML) training has become a necessity with the prevalence of billion to trillion-parameter-scale models.
By Zechen Ma, Zixi Qu, Jinyan Yi, David Lin, Yashar Ganjali
arXiv:2602.14049v2 Announce Type: replace-cross
Abstract: Spatio-temporal traffic forecasting is a core component of intelligent transportation systems, supporting various downstream tasks such as si...
By Yue Wang, Areg Karapetyan, Djellel Difallah, Samer Madanat
FedQoS is a federated learning framework that predicts future QoS failure probabilities for candidate access links in dynamic indoor‑outdoor environments, enabling reliable access‑node selection without centralizing user data. Each access node trains locally on its network logs, while a global QoS‑risk predictor is built through federated aggregation. Simulations using physics‑based synthetic datasets show that FedQoS reduces QoS‑failure rates compared to signal‑based and historical‑QoS heuristics, achieving near‑centralized performance even under non‑IID data conditions.
By Nguyen Van Thieu, Ti Ti Nguyen, Ons Aouedi, Zerihun Huruy, Vu Nguyen Ha, Symeon Chatzinotas
The paper introduces MuViS-C, a multi‑domain benchmark that evaluates the robustness of learning‑based virtual sensing models against ten common sensor failure modes, ranging from subtle drifts to catastrophic dropouts. It assesses models using average error, relative degradation, and worst‑case fragility across nine datasets from six domains, comparing six architectures (gradient‑boosted trees, convolution, recurrence, attention, and MLP‑mixing). The study finds that all models degrade under corruption, gradient‑boosted trees are most robust, and targeted robustification can improve attention models at the cost of nominal performance.
By Jens U. Brandt, Noah C. Puetz, Alexander Windmann, Marc Hilbert, Elena Raponi, Thomas B\"ack, Thomas Bartz-Beielstein