An Intelligent Decision-Making Framework for Secure IoT Data Transmission Using Hybrid Quantum Key Distribution and Deep Neural Networks
Abstract
Quantum cryptography has been developed as a capability to manage the rising risks to security of networks, with data breaches and cyber-attacks. Algorithmic solutions that permit improved selection of quantum key distribution for encryption and authentication were developed. In IoT, the conventional encryption techniques employed are susceptible to numerous threats, and the arrival of quantum computing intensifies these exposures. In this manuscript, we offer an Enhancing Secure Data Transmission Using Hybrid Quantum Key Distribution and Deep Neural Network (SDTHQKD-DNN) algorithm. The SDTHQKD-DNN approach proposes to offer a promising and scalable solution for secure data transmission and intelligent encryption processing in IoT networks. The main novelty of the proposed framework lies in the combination of hybrid Quantum Key Distribution with deep neural network based intelligent classification for simultaneously enhancing data processing and communication security capability in IoT networks. It combines DBN-based classification, multi-stage feature selection and secure key distribution within an integrated secure data transmission framework. The data is standardized using StandardScaler to normalize the feature ranges in the IoT-generated data. To perform dimensionality reduction, a combination of recursive feature elimination, SelectKBest, and mutual information is employed. In addition, the deep belief network (DBN) method is deployed to support secure data transmission through accurate classification, with RMSProp optimization enhancing convergence speed and managing sparse gradients effectively. The experimental results of the proposed model are assessed using benchmark datasets and achieved the accuracy of 99.42%, demonstrates the proposed method outperforms the existing methodologies.