An Edge Computing Matching Framework With Guaranteed Quality of Service

Edge computing is a new computing paradigm, which aims at enhancing user experience by bringing computing resources closer to where data is produced by Internet of Things (IoT). Edge services are provided by small data centers located at the edge of the network, called cloudlets. However, IoT users often face strict Quality of Service (QoS) constraints for a proper remote execution of their applications on edge. Each user has specific resource requirements and budget limitations for her IoT application, while each cloudlet offers a limited number and types of resources, each with a specific cost. Therefore, a key challenge is how to efficiently match cloudlets to IoT applications and enable a convenient any-time access to edge computing services considering preferences and incentives of users and cloudlets. In this article, we address this problem by proposing a novel two-sided matching solution for edge services considering QoS requirements in terms of service response time. In addition, we determine dynamic pricing of edge services based on the preferences and incentives of cloudlets, IoT users, and the system. The proposed matching is incentive compatible, individually rational, weakly budget balanced, asymptotically allocative efficient, and computationally efficient. We perform a comprehensive assessment through extensive performance analysis experiments to evaluate our proposed matching and pricing solutions.