Edge computing (EC) is a promising paradigm for providing ultra-low latency experience for IoT applications at the network edge, through pre-caching required services in fixed edge nodes. However, the supply-demand mismatch can arise while meeting the peak period of some specific service requests. The mismatch between capacity provision and user demands can be fatal to the delay-sensitive user requests of emerging IoT applications and will be further exacerbated due to the long service provisioning cycle. To tackle this problem, we propose the mobile-assisted edge computing framework to improve the QoS of fixed edge nodes by exploiting mobile edge nodes. Furthermore, we devise a CRI (Credible, Reciprocal, and Incentive) auction mechanism to stimulate mobile edge nodes to participate in the services for user requests. The advantages of our mobile-assisted edge computing framework include higher task completion rate, profit maximization, and computational efficiency. Meanwhile, the theoretical analysis and experimental results guarantee the desirable economic properties of our CRI auction mechanism.

中文翻译：

---

用于新兴物联网应用的移动辅助边缘计算框架

边缘计算 (EC) 是一种很有前途的范例，它通过在固定边缘节点中预先缓存所需的服务，为网络边缘的物联网应用程序提供超低延迟体验。但是，在满足某些特定服务请求的高峰期时，可能会出现供需不匹配。容量供应和用户需求之间的不匹配对于新兴物联网应用的延迟敏感用户请求可能是致命的，并且由于服务供应周期长而将进一步加剧。为了解决这个问题，我们提出了移动辅助边缘计算框架，通过利用移动边缘节点来提高固定边缘节点的 QoS。此外，我们设计了一种 CRI（可信、互惠和激励）拍卖机制来激励移动边缘节点参与用户请求的服务。我们的移动辅助边缘计算框架的优势包括更高的任务完成率、利润最大化和计算效率。同时，理论分析和实验结果保证了我们的 CRI 拍卖机制具有理想的经济特性。