The increasing complexity of urban traffic networks has highlighted the potential of Multi-Agent Reinforcement Learning (MARL) for Traffic Signal Control (TSC). However, most existing MARL methods assume homogeneous observation and action spaces among agents, ignoring the inherent heterogeneity of real-world intersections in topology and signal phasing, which limits their practical applicability. To address this gap, we propose HAPS-PPO (Heterogeneity-Aware Policy Sharing Proximal Policy Optimization), a novel MARL framework for coordinated signal control in heterogeneous road networks. HAPS-PPO integrates two key mechanisms: an Observation Padding Wrapper (OPW) that standardizes varying observation dimensions, and a Dynamic Multi-Strategy Grouping Learning (DMSGL) mechanism that trains dedicated policy heads for agent groups with distinct action spaces, enabling adequate knowledge sharing while maintaining structural correctness. Comprehensive experiments in a high-fidelity simulation environment based on a real-world road network demonstrate that HAPS-PPO significantly outperforms Fixed-time control and mainstream MARL baselines (e.g., MADQN, FMA2C), reducing average delay time by up to 44.74% and average waiting time by 59.60%. This work provides a scalable and plug-and-play solution for deploying MARL in realistic, heterogeneous traffic networks.

Connected and automated vehicles (CAVs) are emerging as mobile sensing and actuation resources in mixed traffic, but their contribution to urban road-network resilience depends on more than nominal penetration. This structured critical review examines how local CAV sensing and control can be translated into effective controllability and measurable recovery across vehicle, intersection, corridor, and network scales. An auditable pre-submission update retrieved 1601 raw records from the Web of Science Core Collection, ScienceDirect, and IEEE Xplore and yielded 1476 unique auditable candidates after deduplication. This update identified 66 incremental relevant records, which were combined with the earlier 202-study evidence pool to form a 268-record corpus for the final synthesis. The final corpus was organized into core, bridging, and background/methodological evidence and synthesized using an eight-dimensional evidence framework. The synthesis shows that effective controllability depends on CAV spatial availability, platoon organization, HDV response, communication quality, and computation latency. Signal–vehicle coordination can jointly reshape right-of-way and arrival processes, yet most studies still emphasize nominal delay, energy, throughput, or safety rather than disruption-stage service retention and recovery. We therefore synthesize a cross-scale control architecture and a communication-aware control-quality contract linking traffic and V2X states, cooperation mode, safe fallback, and re-synchronization. Service retention, cumulative performance loss, and recovery time are recommended as resilience objectives, while the E0–E4 maturity ladder highlights the remaining gap between simulation performance and deployment-ready evidence.