Searcharxiv⌕ Search

arXiv · 2611.14961

CommJEPA: When Predictive Representation Learning Meets Wireless Intelligence

Abstract

Artificial intelligence (AI) is becoming an integral part of wireless systems, yet much of today's wireless AI remains task-specific and reactive, with models typically trained to map current observations to predefined outputs or reconstruct observed data. While effective for individual tasks, such approaches provide limited support for learning what is predictable, reusable, and transferable in dynamic wireless environments. Joint-Embedding Predictive Architecture (JEPA) offers a different perspective by learning predictive representations in latent space that capture informative structure in the observations. In this article, we introduce Communication JEPA (CommJEPA) as a communication-oriented framework for predictive representation learning. The central idea is bidirectional. JEPA can equip wireless systems with predictive representations for capturing latent channel and environmental dynamics, integrating heterogeneous observations, adapting to changing conditions, and supporting future decisions. Conversely, the physical structure, partial observability, multiple time scales, action-dependent evolution, and resource constraints of wireless systems motivate new communication-aware JEPA designs. We develop this mutual perspective, organize the resulting design space, and discuss representative JEPA-based wireless instantiations for prediction and control. We further highlight open challenges toward scalable, interactive, and predictive wireless intelligence.

Explore related subjects

Keep this discovery

Explore connections, maps & timelines

BibTeXRIS

Feiran You, Hongyang Du, Kang Wei, Jun Li, Abbas Jamalipour. 2026-11-11. CommJEPA: When Predictive Representation Learning Meets Wireless Intelligence. https://searcharxiv.com/abs/2611.14961

Citations retain the original authors and DOI and include this site’s record URL. If this discovery service contributed to your research, you can also cite Searcharxiv by Indiciae: BibTeX · RIS.

KEEP EXPLORING

Related papers

Flux: Optimal Scheduling of Optical Circuit Switches for LLM Training

Optical Circuit Switching (OCS) offers high bandwidth density and energy efficiency for LLM training, but incurs a non-negligible reconfiguration delay. Prior work typically schedules optical circuit switches independently of compute, using aggregate traffic demand to determine which circuits to provision and when. We argue that this separation creates a fundamental inefficiency: reconfigurations that ignore the compute timeline can stall communication, resulting in low circuit utilization and large buffer requirements. In this paper, we present Flux, a scheduler that optimally schedules optical circuit switches based on the structure of the entire workload. Flux remains effective across a wide range of switching speeds by reusing circuits and amortizing reconfiguration delay behind compute and communication. We show that Flux reduces training iteration time by up to $10\times$ and peak NIC buffer requirements by more than three orders of magnitude compared to traditional periodic schedulers.

cs.NI↗

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

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 $ΔR$, 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~$μ$s per packet.

cs.NI↗

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

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.

cs.NI↗