Cloud Vision Optimizer Framework Using Deep Learning for Efficient Resource Management in Distributed Cloud Environments

In cloud computing model, services are provided to the users at demand over high speed internet. Such an internet based computing paradigm is providing resources with no errors on a rental basis. In cloud computing environments, the problem of allocation of resources to the virtual machines is quite a challenge. The problems of virtual machine allocation have received solutions through the usage of the current static, dynamic, and Meta heuristic methodologies. The local best solution was followed by these strategies. It takes slow convergence rate leading to a locally optimal solution versus a constantly optimal one. To make MAJFOA-DCN, this paper uses the combination of Modified Artificial Jellyfish Optimization Algorithm (MAJFOA) and Deep-Cloud Net (DCN). It is used in this technique to issue virtual machines and solves the problem. Performance metrics for the proposed hybrid technique are assessed and studied using the metrics of Energy Consumption (Kwh) (7.16%), Host Utilization, Average Delay Response Time (ms) (22.11 %), SLA Violations (0.92%) and Number of Migrations (17.82%). The MAJFOA provides the extensive data set to the DCN model to improve its accuracy. The global optimality of the hybrid method has a very fast convergence. Such testing results show that the MAJFOA-DCN exceeds any other conventional approaches in both the real workload situations and the simulated ones. This is due to its ability to efficiently balance resource allocation and minimum latency. In addition, the MAJFOA-DCN framework implementation not only enhances operational efficiency, but also leads to more sustainable (in environmental and financial) and cost effective data center environment.