Attack Prediction in Supply Chain Management Systems Using Hybrid Quantum-Classical Generative Advarsarial Network with War Strategy Optimizer

In the recent past, the advancement of the supply chain management SCM systems and integration of internet of things (IoT) devices pose threats, and for this, the ability to implement intrusion detection has become a necessity. The general approaches have some problems such as high false positive counts among negative cases, low accuracy, and computational constraints. Despite vast improvements seen in the results yielded by ML especially Deep learning techniques, the some of the limitations include over-fitting, bad feature selection and high computational complexity amongst others. To overcome these challenges, we introduced the Hybrid Quantum-Classical Generative Adversarial Network with War Strategy Optimizer for intrusion detection in SCM environments denoted as HQCGAN-WSO. Perform experiments on two reference datasets, BoT-IoT and ModBus with data preprocessing on data normalization with min-max normalization and feature selection with WLS. Then the attacks have been categorized using the generative adversarial network that is optimized with the use of hybrid quantum classical War strategy. HQCGAN-WSO model has promising advantages in accuracy, recall, precision, computational complexity, which provides a feasible and available solution for the SCM intrusion detection. The HQCGAN-WSO model outperforms the current approaches, with increased accuracy of 99.8 % of specificity and 48 ms computational time.