Agentic-V2X: Small Language Model Agents for Deadline-Aware V2X Scheduling in 5G/6G Networks
arXiv:2607.04290v1 Announce Type: cross Abstract: Large Language Models (LLMs) are proposed as control interfaces for next-generation networks, but their latency, hallucinations, and lack of control guarantees make them unsuitable for near-real-time packet schedulers, especially in dynamic V2X environments. This paper introduces Agentic-V2X, an architecture where a small, locally deployed language model acts as a periodic non-real-time rApp-inspired policy creator, while a lightweight xApp-like controller executes validated policies at intervals suitable for scheduling. The framework targets d
The proliferation of LLMs and the increasing demand for low-latency, reliable communication in critical applications like V2X environments are driving research into more optimized AI control interfaces for 5G/6G networks.
This development addresses key limitations of large language models for real-time control, suggesting a viable path for integrating AI into critical infrastructure by leveraging smaller, specialized models for policy creation and execution.
The paradigm shifts from monolithic LLM control to a hybrid architecture combining a periodic policy creator with a lightweight real-time executor, enhancing reliability and reducing latency for AI-driven network management.
- · 5G/6G network providers
- · Automotive industry (V2X)
- · Small language model developers
- · Edge computing infrastructure
- · Monolithic LLM-based control systems
- · Traditional network schedulers
More efficient and reliable V2X communication will enable advanced autonomous driving features and smart city infrastructure.
The validated architecture could be adopted beyond V2X, leading to widespread deployment of similar distributed AI control systems across various critical infrastructures.
This approach may foster a new market for highly specialized LLMs and execution frameworks tailored for specific real-time control applications, accelerating the development of truly autonomous systems.
This signal links to a primary source. Continuum Brief monitors and indexes it as part of the live intelligence stream — we do not republish source content.
Read at arXiv cs.AI