AI-driven Data Security Framework Using AES Encryption
Keywords:
Advanced encryption standard, Artificial intelligence, Autonomous vehicles, Cloud storage security, Cybersecurity, Data preprocessing, Dual encryption, Intrusion detection system (IDS), Internet of transportation systems, Key generation automation, Machine learning, Modified particle swarm optimization (MPSO), Multi-agent neural network (MANN), Over-the-air (OTA) updates, Real-time threat detection, Secure data transmission, SteganographyAbstract
In order to improve information safety, the proposed system offers an adaptive and secure encryption solution that integrates Artificial Intelligence (AI) with the Advanced Encryption Standard (AES). The system performs multi-stage text preprocessing such as keyword extraction, semantic analysis, and redundancy removal, followed by intelligent intrusion detection using a MANN-based model, optimized via Modified Particle Swarm Optimization (MPSO) to improve detection accuracy and reduce false positives.
Real-time data analysis enables anomaly detection, vulnerability forecasting, and dynamic adjustment of encryption parameters, ensuring rapid adaptability to emerging threats. The system classifies data as normal or intrusive, rejecting malicious inputs to safeguard integrity. Normal data undergoes dual-layer encryption (RSA + AES) and is further secured through steganography before cloud storage, providing layered defense mechanisms.
Performance evaluations indicate increased intrusion detection accuracy, reduced processing time due to automated preprocessing, and enhanced data confidentiality through dual encryption and secure cloud storage. This is particularly critical in the Autonomous Vehicle (AV) industry, where sensor data, OTA updates, and cloud infrastructure are vulnerable. The research emphasizes secure supply chains, consistent cybersecurity policies, and active security frameworks powered by AI to safeguard against sophisticated cyberattacks.
