A Cross-Layer Value of Information-Based Communication–Control Joint Optimization Algorithm for Connected Vehicle Platoon

Yaqi Xu, Tony Quek

Keywords: Connected Vehicle Platoon, Cross-layer Value of Information, Joint Optimization, Hierarchical Reinforcement Learning, V2X Resource Allocation, Platoon Control.

Abstract:

Connected vehicle platooning is a promising traffic model that has recently attracted attention from academia and industry. The tight coupling between the control and communication systems involved in platoon necessitates their collaborative design. In this paper, a hierarchical joint optimization algorithm based on the Cross-layer Value of Information (C-VoI), named Hierarchical Dual-Clip Proximal Policy Optimization (HDC-PPO) is proposed. The proposed C-VoI metric provides a unified measure that quantifies the value of information by integrating both control and communication performance.
Building upon this metric, the HDC-PPO algorithm performs multi-timescale joint optimization: the control layer enhances the platoon’s string stability by generating optimal acceleration commands, while the communication layer maximizes the C-VoI by performing subchannel selection and transmission power allocation, thereby realizing value-driven communication resource management.
Simulation results demonstrate that compared with state-of-the-art algorithms, the proposed method significantly improves both string stability and packet reception ratio, achieving superior overall communication efficiency and control performance for the platoon system.

Submission Number: 12