Review on QoS Aware Resource Management in Fog Computing Environment
Hemant Kumar Apat, Bibhudatta Sahoo, Prasenjit Maiti, Punyaban Patel · 2020 IEEE International Symposium on Sustainable Energy, Signal Processing and Cyber Security (iSSSC) · 2020
The ever-increasing demand for computing power and the high penetration of the Internet of Things(IoT) in our daily life interactions undoubtedly enhance the quality of life. To fulfill the heterogeneous demand of these IoT applications Quality of service(QoS) is an important performance measurement metric that must be considered. The traditional computing paradigm is not effective to address the latency sensitive applications generated through the IoT devices. To overcome these limitations of traditional computing a novel computing paradigm Fog computing emerges as a viable solution to provide cloudlike services. However moving the computation, networking, and storage resources to the edge of the network near to source do not guarantee to solve all the challenges. Due to the limited storage capacities of fog nodes, the optimal distribution of resources is an important key to enhance the overall performance of the fog computing paradigm. The resources available in the fog infrastructure are constrained and heterogeneous so there is a critical need to optimize the resources in the fog layer to handle and allocate the resources efficiently to the different applications in an IoT network. To manage the resources efficiently we need to study different resource management issues, objectives, and approaches. Different authors classify the resource management policy into different methods like service placement, resource scheduling, resource allocation, and resource provisioning and adopted different heuristic and meta-heuristic strategies to optimize the available resources in fog layers like computation, communication, storage, data, and energy from the end-user. In this paper we discuss the functionality and importance of fog orchestration node in the generic three-tier architecture that is primarily designed for efficient workload distribution like the number of virtual machines required, the number of different resources required, etc. to execute in the dynamic IoT environment. Also, we introduce some challenges in resource management that must be included in order to provide better QoS to clients.