Cloud Off loading Efficiency With Deepsmote And Ant Lion Optimizer Enhanced By Non-Linear Analysis

Abstract

Managing and processing enormous volumes of data depends on cloud computing, which has evolved into a fundamental instrument. Effective cloud offloading methods determine both reduced latency and optimization of computer resources. Deep SMote, a modified Synthetic Minority over-sampling technique, and the Ant Lion Optimizer (ALO) are rising as possible methods for raising cloud offloading efficiency. Particularly in dynamic environments with imbalanced datasets, conventional cloud offloading methods can find it difficult to balance compute load with mizing latency. Current approaches cannot sufficiently solve the challenges given by high-dimensional data and the complex complexity of offloading decisions. This paper proposes a combination approach to increase cloud offloading efficacy by use of the Ant Lion Optimizer and Deep SMote. Deep Smote generates synthetic samples for balancing unbalanced datasets, therefore improving the quality of the input data for optimization. Inspired by nature, the Ant Lion Optimizer develops a metaheuristic leading to optimal offloading methods. Techniques of non-linear analysis enable fit to complex data patterns and aid to enhance the optimization process. The proposed approach clearly surpasses accepted knowledge. Numerical studies show a 23% drop in latency and an increase in offloading efficiency by 19% compared to baseline techniques. Moreover, using the approach increases general system throughput by 15%. These results show how well DeepSMote and ALO coupled with non-linear analysis tackle cloud offloading issues.