Performance improvement of cloud security with parallel anarchies society optimization algorithm for virtual machine selection in cloud computing

Cloud computing (CC) is a future promising computing technology and has proven tremendous exceptional attainable in managing the hardware and software program sources placed at third-party provider vendors. However, security and virtual machines (VMs) selection to a process a request characterize a large venture. The major problem in the usage of the technology is the security concerns which increase the demand for a robust security mechanism to protect the data on the cloud. Hence in order to resolve this issue, an approach was designed for the purpose of enhancing the cloud security as well as the time of execution known as weighted mean-based convolutional neural network with advanced encryption standard (WMCNN-AES). By employing the method of pseudorandom number generators (PRNGs), random keys are generated and the finest keys are produced by using the WMCNN method. In the WMCNN approach, randomly generated keys are taken as input and the approximation concerning the secret key of the input is produced by the hidden layers. For the purpose of encrypting the data, the finest secret keys are generated by the output layer. By employing the advanced encryption standard (AES) algorithm, encryption is executed. Storage of the files concerning the encrypted data is done in the cloud storage system. Optimal selection of VMs performs a significant enhancement of the performance through reducing the execution time of requests (tasks) coming from stakeholders and maximizing utilization of cloud resources. For the purpose of optimizing the virtual machines’ or VMs’ selection, parallel anarchies society optimization (PASO) is employed in the cloud environment to increase the performance and resource utilization. When comparing the experimental results of the proposed system with the existing system, it is observed that the proposed system accomplished an enhanced performance with respect to resource utilization, cost, throughput, service latency as well as execution time.