期刊
IEEE TRANSACTIONS ON COMMUNICATIONS
卷 67, 期 9, 页码 6221-6233出版社
IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
DOI: 10.1109/TCOMM.2019.2920835
关键词
Mobile edge computing; tandem queue; effective capacity; resource allocation; statistical QoS provisioning
资金
- Beijing Natural Science Foundation [L182038]
- National Natural Science Foundation for Distinguished Young Scholar of China [61325006]
- Natural Science Foundation of China [61461136002]
- 111 Project of China [B16006]
In the mobile edge computing (MEC) network, the applications of devices can be offloaded to the MEC server via the wireless link and then processed through the computation resource, to satisfy the computation and latency demand. Thus, a two-stage tandem queue is formed in the MEC network, consisting of the transmission queue and computation processing queue. However, the fluctuating wireless channel environment not only leads to the stochasticity of service in the first transmission queue, but also brings random computation task arrival in the second computation processing queue, which makes it difficult to guarantee the end-to-end quality of service (QoS) requirement. In this paper, we firstly derive the effective capacity of MEC with the two-stage tandem queue. Further, we formulate the joint bandwidth and computation resource allocation problem under the statistical QoS guarantee, to maximize the total revenue of network. This problem is proven to be NP-hard by the reduction to the two-dimensional knapsack problem. Then we propose an efficient algorithm based on alternating direction method of multipliers (ADMM) to reduce the computation complexity, where the complicated problem can be decomposed and transformed into some convex subproblems. Simulation results reveal the inherent relationship between the required bandwidth and computation resource in terms of the supported arrival rate and end-to-end delay, and also demonstrate the proposed scheme can achieve better performance than other schemes.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据