SGCP: A Self-Organized Game-Theoretic Framework For Collaborative Perception

TL;DR

The paper tackles scalable, infrastructure-free collaborative perception for dense urban vehicular networks by introducing a two-stage, fully decentralized framework. It decomposes the problem into a perception-driven coalition game for self-organized cluster formation and a distributed resource scheduling game within clusters, yielding monotonic improvements in a global potential Φ(a). Through a hierarchical fusion architecture—intra-cluster early fusion and inter-cluster late fusion—coupled with a perception-priority scheduling strategy, the method reduces communication overhead while maintaining or improving perception accuracy. Experimental results on the CARLA–OpenCDA–NS3 platform show substantial mAP gains (e.g., mAP@0.5 and mAP@0.7) and lower bandwidth consumption compared to RSU-based and other baselines, demonstrating practical impact for infrastructure-free CP in high-density traffic.

Abstract

Collaborative perception holds great promise for improving safety in autonomous driving, particularly in dense traffic where vehicles can share sensory information to overcome individual blind spots and extend awareness. However, deploying such collaboration at scale remains difficult when communication bandwidth is limited and no roadside infrastructure is available. To overcome these limitations, we introduce a fully decentralized framework that enables vehicles to self organize into cooperative groups using only vehicle to vehicle communication. The approach decomposes the problem into two sequential game theoretic stages. In the first stage, vehicles form stable clusters by evaluating mutual sensing complementarity and motion coherence, and each cluster elects a coordinator. In the second stage, the coordinator guides its members to selectively transmit point cloud segments from perceptually salient regions through a non cooperative potential game, enabling efficient local fusion. Global scene understanding is then achieved by exchanging compact detection messages across clusters rather than raw sensor data. We design distributed algorithms for both stages that guarantee monotonic improvement of the system wide potential function. Comprehensive experiments on the CARLA-OpenCDA-NS3 co-simulation platform show that our method reduces communication overhead while delivering higher perception accuracy and wider effective coverage compared to existing baselines.

Figures (4)

• Figure 1: System overview: CAVs self-organize into clusters, each with a leader. Members upload high-value point clouds to leaders for early fusion. Cavs broadcast detection results for late fusion across clusters.
• Figure 2: Decentralized collaborative perception framework.
• Figure 3: A complex urban intersection scenario in CARLA Town03.
• Figure 4: System communication overhead (Mbps). Lower values indicate higher communication efficiency.