A container architecture for migrating interoperable applications using weight adaptive temporal CNN and secretary bird optimization algorithm in cloud computing

Virtualization in cloud computing enables multiple virtual servers to be deployed on a single physical infrastructure. The increasing number of cloud applications also increases resource overload during service execution. However, the main challenge with the migration of interoperable applications is the process of establishing optimal resource allocation, performance, quality of service, the tradeoff between cost-effectiveness and satisfying the requirements of service level agreement. Deep learning methods are useful for load balancing because they can identify noisy workload fluctuations. In this paper, Weight Adaptive Temporal CNN and Secretary Bird Optimization Algorithm (WATCSBOA)-based container architecture is proposed for predicting the workload, which optimizes the degree of resource utilization and reduces the overload during the process of executing the migrated interoperable applications. This container architecture uses Weight Adaptive Temporal CNN (WATCNN) to determine long-term sequence dependencies for deciding the actual workload thereby improving the accuracy and response speed of container clusters. It mainly helps enhance the capabilities of feature extraction when the deep learning model is employed to address long sequences. It uses the Secretary Bird Optimization Algorithm (SBOA), which helps in parameter optimization to maximize resource allocation and attain optimized migration of interoperable applications. It supports resource allocation and migration by evaluating a fitness function that considers demand, transmission cost, load prediction, and resource availability. The simulation results show that the proposed WATCSBOA scheme improves load prediction performance by reducing RMSE by 11.26%, MAE by 9.64%, and MAPE by 8.96%, while improving PCC by 12.36% compared with the baseline approaches. Under the moderate and high workload scenarios, WATCSBOA reduces migration time by 11.64% and 10.98%, respectively, and improves resource utilization by 12.39% and 11.64%, respectively.