Optimizing VM Allocation with Queue Dependent Requests in fog Network

Abstract

The volume of data is continuously expanding as a result of the quick rise of the Industrial Internet of Things (IIoT), social media, digitization, as well as wireless communication in numerous sectors. With the help of fog computing, cloud computing is an emerging technique for handling and analyzing enormous amounts of data storage. Fog Computing is a set of methods for improving the quality of services (QoS) offered to consumers via cloud computing, which is becoming increasingly overburdened as a result of enormous data flows. All of the data is being sent to the cloud, and retrieving it from there causes a lot of latency and necessitates a lot of network capacity. The fog nodes are seen as a heterogeneous multi-VM system with a finite queue in which the VMs are shared by many client requests. When the system's request queue length exceeds a threshold value Nj (j = 1,2,….r-1), the (j + 1)thVM begins processing the requests and continues until the waiting buffer length again reduced to the same level. The recursive technique is used to obtain the steady-state queueing size distribution, which takes into account Markovian arrival with service time. We derived several system properties and studied the fog system's performance based on the client requests and the queue length.