期刊
IEEE JOURNAL ON SELECTED AREAS IN COMMUNICATIONS
卷 38, 期 5, 页码 803-815出版社
IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
DOI: 10.1109/JSAC.2020.2980908
关键词
Fog computing; industrial Internet of Things; non-orthogonal multiple access; delay-energy tradeoff; online learning
资金
- Natural Science Foundation of China (NSFC) [61932014]
- Nature Science Foundation of Shanghai [19ZR1433900]
- Huawei Technologies [YBN2019125163]
Fog computing (FC) has the potential to process computation-intensive tasks in Industrial Internet of Things (IIoT) systems. In parallel with the development of FC, non-orthogonal multiple access (NOMA) has been recognized as a promising technique to significantly improve the spectrum efficiency. In this paper, a NOMA-based FC framework for IIoT systems is considered, where multiple task nodes offload their tasks via NOMA to multiple nearby helper nodes for execution. We formulate a joint task scheduling and subcarrier allocation problem, with an objective to minimize the total cost in terms of the delay and energy consumption, while taking into account the practical communication and computation constraints. Note that the task scheduling includes task, computation resource, and power allocations. Since the task and subcarrier allocations involve binary variables, it is challenging to obtain an optimal solution for such a combinatorial problem. To this end, we solve the task scheduling and subcarrier allocation problem in an online learning fashion. During the online learning process, we propose an iterative algorithm to jointly optimize the subcarrier allocation and task scheduling in each time episode. Simulation results show that the proposed scheme can significantly reduce the sum cost compared to the baseline 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