LLM-Empowered Cooperative Content Caching in Vehicular Fog Caching-Assisted Platoon Networks

TL;DR

This work tackles content caching latency in VFC-assisted platoon networks by introducing a three-tier cooperative caching architecture that leverages LLMs for real-time decisions. It fuses heterogeneous data into a hierarchical prompt ${\mathcal{P}^r = \text{Role} \oplus D_{\text{task}} \oplus \mathcal{H}^r}$ and derives a two-stage process where the LLM outputs an ordered list $L^r$ that a deterministic mapper converts into caching decisions $X^r$. Through a two-tier optimization, the approach enforces capacity constraints and content exclusivity with $\sum_{j=1}^N x_{j,f}^r + \sum_{k=1}^{N^r} y_{k,f}^r + z_f^r = 1$, aiming to minimize the average retrieval delay across users and contents. Simulations show improved Average Cache Hit Ratio (ACHR) and reduced Average Content Transmission Delay (ACTD) relative to baselines, with the biggest gains from expanding platoon cache capacity; future work includes multi-platoon collaboration and security considerations.

Abstract

This letter proposes a novel three-tier content caching architecture for Vehicular Fog Caching (VFC)-assisted platoon, where the VFC is formed by the vehicles driving near the platoon. The system strategically coordinates storage across local platoon vehicles, dynamic VFC clusters, and cloud server (CS) to minimize content retrieval latency. To efficiently manage distributed storage, we integrate large language models (LLMs) for real-time and intelligent caching decisions. The proposed approach leverages LLMs' ability to process heterogeneous information, including user profiles, historical data, content characteristics, and dynamic system states. Through a designed prompting framework encoding task objectives and caching constraints, the LLMs formulate caching as a decision-making task, and our hierarchical deterministic caching mapping strategy enables adaptive requests prediction and precise content placement across three tiers without frequent retraining. Simulation results demonstrate the advantages of our proposed caching scheme.

Figures (2)

• Figure 1: System scenario
• Figure 2: Performance comparison of caching schemes