SPAMURAI: An Intelligent Quantum-Secure SMS Filtering and Encryption Framework Using Post-Quantum Cryptography

Abstract

With the growing dependence on Short Message Service (SMS) for sensitive applications such as banking alerts and user authentication, ensuring both message authenticity and communication security has become a critical challenge. Traditional Machine Learning (ML)-based spam detection systems are effective in identifying malicious content but do not secure the transmission process. Additionally, the emergence of quantum computing threatens classical cryptographic schemes, necessitating the adoption of quantum-resistant solutions. The system utilizes a Random Forest classifier with TF-IDF feature extraction to accurately classify SMS messages. To secure communication, the framework incorporates CRYSTALS-Kyber for key encapsulation, which enables efficient and secure session key generation based on the hardness of lattice problems such as Learning With Errors (LWE). Additionally, CRYSTALSDilithium is implemented to provide strong authentication and message integrity through quantum-resistant digital signatures. The proposed system follows a layered architecture comprising secure user authentication, real-time message classification, PQCbased encryption, and protected data storage. The integration of Kyber and Dilithium ensures resistance against both classical and quantum adversaries while maintaining computational efficiency suitable for mobile environments. Experimental evaluation demonstrates that the system achieves reliable spam detection performance along with enhanced end-to-end security. This implementation highlights a practical and scalable approach for developing next-generation secure messaging systems capable of addressing both present-day cyber threats and future quantum risks.