An Empirical Study on Various Workflow Scheduling Techniques in Cloud Computing

Cloud computing (CC) is a modern technology which deploys networks of servers, positioned in extensive remote areas, for conducting operations on a larger quantity of data. In CC, a workflow module is employed for representing diverse scientific and web applications. One of the major problems is scheduling larger workflows of tasks on the heterogeneous CC environment. This article summarizes some important research works in the field of CC associated with workflow scheduling (WS). At first, this survey represents the common structure of WS in cloud and then, scheduling approaches are catalogued into four techniques like Machine Learning (ML), Deep Learning (DL), Hybrid and Algorithmic-based approaches are discussed in detail. Thereafter, the research gaps faced by the traditional approaches for WS will help the researchers in future for building an innovative model. This study considers 25 research articles focused on different methodologies employed for WS in cloud. Lastly, the evaluation is performed in the survey regarding year of publication, performance measures, research technique, and attainment of the research models. It is acknowledged that algorithmic based modules are usually employed approaches, while cloudsim is repeatedly applied tool for WS in CC. Moreover, makespan is the utilized analytic measure with superior value.