An Augmented Salp-swarm Optimization Based on Paillier Federated Multi-layer Perceptron (Pf-mlp) and Homomorphic Encryption Standard (Hes) Techniques for Data Security in Cloud Systems

Providing security to cloud data is one of the essential problems that have needed to be addressed in recent times due to the advancement and development of security breaches in technologies. As a result, the majority of existing research efforts aim to develop various types of cryptographic techniques for ensuring the data security of cloud systems. However, it faced challenges with complex computational operations, inefficient security models, high time consumption, and error outputs. Therefore, the proposed work aims to develop an advanced and hybrid optimization-based cryptographic methodology for increasing the security of cloud data. An Improved Salp-Swarm Optimization (ISSO) technique is deployed to obtain the random number required for secret key generation. In this work, two different encryption techniques, such as the Homomorphic Encryption Standard (HES) and the Paillier Federated Multi-Layer Perceptron (PF-MLP) model, are used for strengthening the security of the original health tweet dataset. Here, the efficacy and security level of these two encryption methodologies are validated for the purpose of identifying the most suitable mechanism to secure the health tweet dataset. For these approaches, the key pair is optimally generated by using an ISSO technique. This type of key generation can complicate matters for users who attempt to attack the original information. Moreover, the novel contribution of this work is that it incorporates the functions of advanced optimization and cryptographic mechanisms for securing the health tweet dataset against attacking users. For validating the performance of this system, various evaluation metrics have been used, and the validated results are compared with the proposed system to demonstrate the improvement of the proposed system.