Secure and Ultra-Reliable 6G IoT Algorithms with AI-Enhanced Shared Key Physical Layer Security

This paper presents a new class of precoder-based Physical Layer Security (PLS) algorithms for 6G mobile networks, a significant advancement that combines security and ultra-reliable low latency communications (S-URLLC). The paper also explores the potential of S-URLLC algorithms for realtime shared secret key information exchange in Layer-1 transmission and proposes AI/ML-based detection algorithms for natively securing future 6G-IoT networks against remote eavesdropping. The paper marries a temporal-spatial M-dimensional code with operator matrices drawn from the class of random unitary (precoder) operators to jointly guarantee ultra-reliability and security. The code delivers ultra-reliability, while the precoder operator safeguards the signal waveform from eavesdroppers. The shared secret key-based model operating within a Hyper-Dimensional Temporal-Spatial Precoding with ML-based decoding (HD-TSP-ML) scheme scales across multi-antenna array sizes and codebook modulation orders and delivers impressive 1215 dB improvements in link reliability. The security analysis of Bit Error Rate (BER) versus Signal-to-Noise Ratio (SNR) of intended receiver versus eavesdropper reveals a low probability of interception by malicious eavesdroppers, even when the passive eavesdropper observes in a high SNR, correlated channel with the intended recipient. The paper presents a system architecture of the shared secret key HD-TSP-ML model, discusses simulation result evaluation of the BER of secret information exchanged for the algorithm in Rayleigh fading plus noise and analyzes numerical model of the transmit data-rate versus MIMO array size as well as latency for the system along with a complexity analysis.