Show simple item record

AuthorZhang, Sunxuan
AuthorLiao, Haijun
AuthorZhou, Zhenyu
AuthorWang, Yang
AuthorZhang, Hui
AuthorWang, Xiaoyan
AuthorMumtaz, Shahid
AuthorGuizani, Mohsen
Available date2022-11-10T06:51:57Z
Publication Date2021-01-01
Publication Name2021 IEEE Global Communications Conference, GLOBECOM 2021 - Proceedings
Identifierhttp://dx.doi.org/10.1109/GLOBECOM46510.2021.9686001
CitationZhang, S., Liao, H., Zhou, Z., Wang, Y., Zhang, H., Wang, X., ... & Guizani, M. (2021, December). Federated Deep Actor-Critic-Based Task Offloading in Air-Ground Electricity IoT. In 2021 IEEE Global Communications Conference (GLOBECOM) (pp. 1-6). IEEE.‏
ISBN9781728181042
URIhttps://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85127251717&origin=inward
URIhttp://hdl.handle.net/10576/36058
AbstractThe integration of air-ground electricity internet of things (AGE-IoT) and machine learning, enables flexible network coverage and intelligent task offloading. However, dynamics of AGE-IoT networks, incomplete information, and resource allocation coupling are still major challenges in achieving intelligent AGE-IoT. In this paper, we investigate a joint multi-timescale task offloading and power control optimization problem to minimize the queuing delay of all the EIoT devices under the long-term constraint of energy consumption. We firstly decompose the joint optimization problem and transform it to large-timescale task offloading optimization and small-timescale power control optimization. Then, we propose a fed-erated deep actor-critic-based task offloading algorithm (FDAC) with two actor-critic networks for multi-timescale optimization. Numerical results show that FDAC has excellent performances in queuing delay and energy consumption compared with existing algorithms.
SponsorThis work was supported by the Science and Technology Project of State Grid Corporation of China under Grant Number SGSDDK00KJJS1900405.
Languageen
PublisherInstitute of Electrical and Electronics Engineers Inc.
SubjectTask Offloading
TitleFederated Deep Actor-Critic-Based Task Offloading in Air-Ground Electricity IoT
TypeConference Paper
dc.accessType Abstract Only


Files in this item

FilesSizeFormatView

There are no files associated with this item.

This item appears in the following Collection(s)

Show simple item record