Usage Aware VNF Placement for Improved QoS in Edge Computing
Briytone Mutichiro, Hyunsik Yang, Younghan Kim · 2019
The advent of edge computing has given rise to new opportunities and challenges in the domain of cloud computing. While the key deliverable is to ensure reliable service provision to users, the characteristics of the edge cloud including high latency, unreliable bandwidth, real-time response, distance from the edge cloud among others, make it challenging to meet set targets. A major challenge operators face is the optimal placement of virtual network functions (VNFs). This problem has been widely researched in the literature of data centers from an operator's perspective little attention given to the user. We formulate the usage-aware placement problem for enhanced user quality of service (QoS) in edge environment. We aim to minimize the user response time through efficient placement and model the user requests in terms of user preference and service selection probability in influencing placement of VNFs in edge-cloud environments. Further we formulate a heuristic usage-aware placement algorithm solution for placement of these service requests. Our results show that our approach achieves considerable reduction in user time response and minimizes resource costs. It meets the QoS requirement for latency sensitive services.