An Intelligent Cloud Computing Security Approach Using a Dual Hybrid Cipher System

Abstract

Although the development of cloud computing has led to a paradigm shift in the data storage, it has also subjected the sensitive information to massive security and performance threats; that is, the data integrity and processing overhead. There is a basic trade-off in classical cryptography: the symmetric-key protocols are really fast, whereas they do not provide secure exchange of keys, but the asymmetric-key protocols are very secure and computationally infeasible when dealing with large amounts of data. Conventional cryptographic techniques commonly have a fundamental dilemma: symmetric-key protocols can offer fast execution but do not include any secure key-exchange protocols, whereas asymmetric-key systems can offer high protection but cannot be used with large volumes of data due to their slow computation speed. To alleviate these problems, this study presents a streamlined dual-cipher architecture that is aimed at boosting the efficiency of the cloud through the integration of the AES-256 and RSA-2048. The intelligence of the system lies in a dynamic workload distribution mechanism such that the high-resource encapsulation of the RSA operations is limited to the encryption of keys, while the bulk encryption is left to the high-speed AES engine. The empirical assessment of different file sizes (1 MB to 1 GB) shows that a sustainable throughput of 56-72 MB/s is achieved at 83-86 percent of the functional speed of standalone AES. As a result, the suggested model provides a viable and scalable design for modern cloud infrastructures with a very low overhead of just 20-21 percent or so relative to single deployments of RSA.