Post-Quantum ZKP for Privacy-Preserving Authentication and Model Verification in Decentralized CAV

Decentralized and Connected Autonomous Vehicle (CAV) networks offer promising advances in safety, efficiency, and real-time decision-making. However, they face significant challenges in authentication, privacy, and scalability—especially in the face of quantum adversaries. This paper proposes a novel post-quantum secure framework that integrates lattice-based Zero-Knowledge Proofs (ZKPs), optimized Binius proofs, and Multi-Layer Compressed Counting Bloom Filters (ML-CCBF) to enable privacy-preserving authentication and model verification in decentralized CAV environments. Our lattice-based ZKP scheme achieves cryptographic commitments in under 5μs, while Binius proofs verify model integrity in less than 0.17 seconds per update. ML-CCBF ensures scalable membership filtering with 0% false positives across 1000 nodes. Experimental results confirm 100% ZKP soundness, strong resilience against simulated quantum and adaptive attacks, and stable latency under increasing network load. These findings demonstrate that our framework delivers quantum-resilient security, real-time efficiency, and robust scalability, offering a viable solution for trustworthy decentralized intelligence in next-generation vehicular systems.