Machine learning techniques in emerging cloud computing integrated paradigms: A survey and taxonomy

Abstract

Cloud computing offers Infrastructure as a Service (IaaS), Platform as a Service (PaaS), and Software as a Service (SaaS) to provide compute, network, and storage capabilities to the clients utilizing the pay-per-use model. On the other hand, Machine Learning (ML) based techniques are playing a major role in effective utilization of the computing resources and offering Quality of Service (QoS). Based on the customer’s application requirements, several cloud computing-based paradigms i.e., edge computing, fog computing, mist computing, Internet of Things (IoT), Software-Defined Networking (SDN), cybertwin, and industry 4.0 have been evolved. These paradigms collaborate to offer customer-centric services with the backend of cloud server/data center. In brief, cloud computing has been emerged with respect to the above-mentioned paradigms to enhance the Quality of Experience (QoE) for the users. In particular, ML techniques are the motivating factor to backend the cloud for emerging paradigms, and ML techniques are essentially enhancing the usages of these paradigms by solving several problems of scheduling, resource provisioning, resource allocation, load balancing, Virtual Machine (VM) migration, offloading, VM mapping, energy optimization, workload prediction, device monitoring, etc. However, a comprehensive survey focusing on multi-paradigm integrated architectures, technical and analytical aspects of these paradigms, and the role of ML techniques in emerging cloud computing paradigms are still missing, and this domain needs to be explored. To the best of the authors’ knowledge, this is the first survey that investigates the emerging cloud computing paradigms integration considering the most dominating problem-solving technology i.e., ML. This survey article provides a comprehensive summary and structured layout for the vast research on ML techniques in the emerging cloud computing paradigm. This research presents a detailed literature review of emerging cloud computing paradigms: cloud, edge, fog, mist, IoT, SDN, cybertwin, and industry 4.0 (IIoT) along with their integration using ML. To carry out this study, majorly, the last five years (2017-21) articles are explored and analyzed thoroughly to understand the emerging integrated architectures, the comparative study on several attributes, and recent trends. Based on this, research gaps, challenges, and future trends are revealed.