Enhanced Quantum Key Distribution and Error Minimization for 6G Networks through Optimized Multi-Head Self-Attention and Gated-Dilated Convolutional Neural Networks

Abstract

6G frameworks will be able to utilize higher frequencies than 5G frameworks and provide altogether higher capacity and much lower latency. First, 6G network input signals stand up to channel obstacles to open security. QKD, a cryptographic technique, allows two parties to safely exchange secret keys for message encryption and decryption. Based on quantum mechanics, any threats in the system warn the authorized users. Qubits can be distributed over fibre or free space quantum channels and encoded with several degrees of freedom. In this proposed method, QKD and Error Rate Optimization using an optimized Multi-Head Self-Attention and Gated-Dilated Convolutional Neural Network (QKD-ERO-MSGCNN) are used to extend the maximum transmission distance and improve the secret key rate (SKR). The variable velocity strategy particle swarm optimization algorithm (VVSPSOA) is used on the input signals to improve the SKR and is fed to the MSGCNN to analyse the quantum bit error rate (QBER) reduction effectively. MSGCNN alone does not optimize the QBER reduction in the channel accurately; intensified sand cat swarm optimization (ISCSO) is used to analyse the QBER reduction accurately. The proposed QKD-ERO-MSGCNN is implemented, and metrics such as accuracy, precision, SKR and QBER reduction attain higher performances when compared with existing techniques.