Optimized Artificial Neural Network-based Approach for Task Scheduling in Cloud Computing

Abstract

Cloud computing has transformed how businesses utilize computing resources. The fundamental aim remains to convene distributed resources such that they operate most efficiently and thus improve overall throughput. This shift is proving useful in developing complex computations on a large scale in commercial practice. The best feature of commercial clouds is their elasticity, which recommends dynamic resource allocation adjustments by users accordingly to real-time demand. Alongside the pay-as-you-go cost model, this ensures that organizations optimise usage and costs. However, good allocation of resources is one of the biggest challenges, known as job scheduling. This refers to how to optimally assign end-user requests to cloud resources, based on the premise that each task can run as quickly as possible. A job scheduler's main objective is to select the best resource to satisfy a user's task, taking into consideration statistical data and dynamic parameters of user jobs. For this problem, various techniques have been researched by the investigators, with AI playing a major role. Among these techniques, the use of genetic algorithms and ant colony optimization techniques is employed to allocate resources optimally, improving the efficiency of job scheduling in the cloud. This research is geared towards continually improving job scheduling in cloud computing. The introduction of artificial neural networks represents an exciting avenue for enhancing optimization and addressing the ongoing challenges faced in efficiently allocating cloud resources.