Enhancing Cloud-Based Inventory Management: A Hybrid Blockchain Approach With Generative Adversarial Network and Elliptic Curve Diffie Helman Techniques

Abstract

In contemporary business organizations, the pivotal role of automation in control processes is evident through Inventory Management Systems (IMS), which leverage advanced techniques and data analytics algorithms to optimize inventory levels, enhance accuracy, and minimize costs. However, existing security techniques for Inventory Management Systems (IMS), including access controls, firewalls, and audits, face challenges in effectively addressing the evolving threat landscape. These limitations, including struggles with dynamic user roles, susceptibility to data manipulation, and challenges in thwarting various cyber threats, necessitate innovative solutions for robust real-time management and security. Consequently, this work proposes a novel hybrid approach that integrates blockchain with RFID data, Generative Adversarial Networks, and Elliptic Curve Diffie-Hellman cryptographic techniques. In the developed hybrid approach, RFID readers are leveraged to collect inventory data, while the Generative Adversarial Network is specifically designed for processing the raw dataset, encompasses data filtering, normalization, and error correction tasks. The utilization of the Elliptic Curve Diffie-Hellman technique is integral for generating both private and shared keys, facilitating secure transmission between the IMS client and cloud-based servers. The blockchain module is engineered to enhance data security and protect shared secret keys, which is achieved through a two-layer mechanism involving encryption via the Advanced Encryption Standard algorithm and SHA-256 hashing function. Additionally, it incorporates the Artificial Algae Algorithm and an Elman Neural Network to ensure robust data access and integrity. To assess the effectiveness of the proposed hybrid approach, it is implemented on a publicly available dataset. The performance assessment involves a comprehensive comparison with state-of-the-art security methods, considering key metrics such as encryption time, decryption time, key generation time, throughput, latency, and data confidentiality rate. Simulation results conclusively demonstrate that the proposed hybrid approach significantly reduces encryption time, decryption time, key generation time, and latency. Furthermore, it notably improves throughput and data confidentiality rates while aligning with stringent IMS security requirements.