Technology Category
Planning, Routing, and Scheduling: Optimization of Spatio-temporal SystemsIntelligent Robots: State EstimationMultiagent Systems: Agent Communication
Connectivity Management in Satellite-Aided Vehicular Networks with Multi-Head Attention-Based State Estimation
🤖 AI Summary
To address connection management challenges in 6G integrated space-air-ground vehicular networks—arising from dynamic topologies and partial observability—this paper proposes a multi-agent collaborative decision-making framework. Methodologically, it integrates multi-head self-attention to enhance robust state estimation under partial observability; employs self-imitation learning to accelerate policy convergence; and leverages channel fingerprinting for low-overhead, high-accuracy link-state awareness. The framework enables autonomous, joint optimization of handovers across heterogeneous V2S, V2I, and V2V links. Evaluated via SUMO-based simulations compliant with 3GPP standards, the approach achieves up to a 14% improvement in transmission utility while maintaining over 92% state estimation accuracy across varying vehicle densities and levels of inter-agent information sharing. These results demonstrate significant gains in real-time responsiveness and environmental adaptability within highly dynamic network conditions.
Technology Category
Application Category
📝 Abstract
Managing connectivity in integrated satellite-terrestrial vehicular networks is critical for 6G, yet is challenged by dynamic conditions and partial observability. This letter introduces the Multi-Agent Actor-Critic with Satellite-Aided Multi-head self-attention (MAAC-SAM), a novel multi-agent reinforcement learning framework that enables vehicles to autonomously manage connectivity across Vehicle-to-Satellite (V2S), Vehicle-to-Infrastructure (V2I), and Vehicle-to-Vehicle (V2V) links. Our key innovation is the integration of a multi-head attention mechanism, which allows for robust state estimation even with fluctuating and limited information sharing among vehicles. The framework further leverages self-imitation learning (SIL) and fingerprinting to improve learning efficiency and real-time decisions. Simulation results, based on realistic SUMO traffic models and 3GPP-compliant configurations, demonstrate that MAAC-SAM outperforms state-of-the-art terrestrial and satellite-assisted baselines by up to 14% in transmission utility and maintains high estimation accuracy across varying vehicle densities and sharing levels.
Problem
Research questions and friction points this paper is trying to address.
Managing dynamic connectivity in satellite-vehicular networks for 6G
Overcoming partial observability in multi-link V2S/V2I/V2V communication
Enhancing state estimation with multi-head attention under limited information
Innovation
Methods, ideas, or system contributions that make the work stand out.
Multi-head attention for state estimation
Multi-agent reinforcement learning framework
Self-imitation learning and fingerprinting
🔎 Similar Papers
No similar papers found.
💼 Related Jobs
No related jobs found.
