It is increasingly popular that platforms integrate various services into mobile applications due to the high usage and convenience of mobile devices, many of which demand high computational capacities and energy, such as cryptocurrency services based on blockchain. However, it is hard for mobile devices to run these services due to the limited storage and computational capacity. In this paper, the problem of computation offloading that requires sufficient computing resources with high utilization in large-scale users and multiprovider MEC system was investigated. A mechanism based on the combinatorial double auction, G-TRAP, is proposed in this paper to solve the above problem. In the mechanism, resources are provided both in the cloud and at the edge of the network. Mobile users compete for these resources to offload computing tasks by the rule that the edge-level resources will be allocated at first while cloud-level resources could be the supplement for the edge level. Given that the proof-of-work (PoW), the core issue of blockchain application, is resource-expensive to implement in mobile devices, we provide resource allocation service to users of blockchain application as experimental subjects. Simulation results show that the proposed mechanism for serving large-scale users in a short execution time outperforms two existing algorithms in terms of social utility and resource utilization. Consequently, our proposed system can effectively solve the intensive computation offloading problem of mobile blockchain applications.

中文翻译：

---

移动拍卖中基于双拍卖的两级资源分流机制

由于移动设备的高使用率和便利性，平台将各种服务集成到移动应用程序中变得越来越流行，其中许多移动设备需要高计算能力和精力，例如基于区块链的加密货币服务。但是，由于有限的存储和计算能力，移动设备很难运行这些服务。本文研究了在大型用户和多提供商MEC系统中需要大量计算资源，高利用率的计算分流问题。为解决上述问题，本文提出了一种基于组合式双拍卖的机制G-TRAP。在该机制中，在云和网络边缘都提供了资源。移动用户通过首先分配边缘级资源，而云级资源可以作为边缘级的补充，从而竞争这些资源来减轻计算任务的负担。鉴于工作量证明（PoW）是区块链应用程序的核心问题，要在移动设备中实施它会消耗大量资源，因此我们向作为实验对象的区块链应用程序用户提供资源分配服务。仿真结果表明，所提出的在短时间内为大型用户服务的机制在社会效用和资源利用方面均优于现有的两种算法。因此，我们提出的系统可以有效地解决移动区块链应用程序的密集计算卸载问题。