Post Quantum Blockchain Trust Framework Using Quantum Support Vector Machines and Deep RL with Chaotic Aquila Optimization

The increasing threat posed by quantum computing to classical cryptographic primitives necessitates the development of resilient blockchain architectures. This paper introduces a Post-Quantum Blockchain Trust Framework that integrates Quantum Support Vector Machines (QSVM) for quantum-secure trust classification, Deep Reinforcement Learning (DRL) for adaptive consensus optimization, and Chaotic Aquila Optimization (CAO) for dynamic hyperparameter tuning. The QSVM module employs quantum kernel mapping to distinguish malicious nodes, while the DRL agent autonomously adjusts transaction validation and propagation strategies based on network feedback. CAO further enhances convergence speed and prevents premature stagnation through chaotic search dynamics, ensuring optimal performance under adversarial conditions. The framework is formulated within a hybrid quantum-classical architecture, incorporating trust score optimization, entropy-driven policy learning, and quantum state encoding for enhanced robustness. Experimental evaluation using the QDataSet demonstrates that the proposed framework outperforms post-quantum baselines, achieving a peak trust classification accuracy of 92.33%, minimum transaction latency of 74.8 ms, and a Quantum Attack Resilience Index of 0.93. Furthermore, the architecture requires only 13 qubits and attains an average gate fidelity of 0.98, confirming its viability on near-term quantum devices. These results validate the proposed method as a scalable, secure, and latency efficient solution for next-generation post-quantum blockchain ecosystems.