TY - GEN
T1 - Two-Tier Multi-Access Partial Computation Offloading via NOMA
T2 - 31st Wireless and Optical Communications Conference, WOCC 2022
AU - Li, Yang
AU - Wu, Yuan
AU - Bi, Suzhi
AU - Qian, Liping
AU - Quek, Tony Q.S.
AU - Xu, Chengzhong
AU - Shi, Zhiguo
N1 - Funding Information:
ACKNOWLEDGMENT This work was supported in part by Science and Technology Development Fund of Macau SAR under Grants 0015/2019/AKP, 0060/2019/A1, and 0162/2019/A3, in part by FDCT-MOST Joint Project under Grant 0066/2019/AMJ, in part by National Natural Science Foundation of China under Grants 62122069, 62072490, 62071431 and 61871271, in part by the Guangdong Basic and Applied Basic Research Foundation under Grant 2022A1515011287, in part by the Key Project of Department of Education of Guangdong Province under Grant 2020ZDZX3050, in part by the Shenzhen Science and Technology Program under Project JCYJ20210324093011030, and in part by the National Research Foundation, Singapore, and Infocomm Media Development Authority under its Future Communications Research & Development Programme.
Publisher Copyright:
© 2022 IEEE.
PY - 2022
Y1 - 2022
N2 - Multi-access edge computing has been considered as a promising solution for enabling computation-intensive yet latency-sensitive applications at resource-constrained wireless devices (WDs). To improve the spectrum efficiency for multi-WD computation offloading, this paper considers non-orthogonal multiple access (NOMA) assisted two-tier multi-access edge computing scenario, which exploits the computation resources of both the edge servers (ESs) and the cloudlet server (CS) deployed at different tiers. In particular, the WDs can offload partial workloads to different ESs simultaneously via NOMA, and the ESs can form a NOMA-group to further offload partially received workloads to the CS for processing. We investigate the total energy consumption minimization problem by jointly optimizing the two-tier offloading decisions, the NOMA transmission duration, and the computation resource allocation. Due to the successive interference cancellation in the NOMA and the coupling effect in two-tier offloading, the formulated optimization problem is strictly non-convex. To address this difficulty, we exploit the hierarchical relationship among the joint optimization variables, and then propose a hybrid deep reinforcement learning (HDRL) algorithm to learn two policies that determine the coupled variables, i.e., the ESs' offloading decisions and the NOMA transmission duration, respectively. Then, the remaining decision variables can be jointly optimized by using the convex optimization methods directly based on the results provided by the HDRL algorithm. Specifically, the HDRL algorithm that uses different policies to determine the coupled variables can converge faster than the existing solutions that learn a single policy to determine all variables. Experimental results are provided to validate the performance of our proposed HDRL algorithm in comparison with two other learning-based algorithms.
AB - Multi-access edge computing has been considered as a promising solution for enabling computation-intensive yet latency-sensitive applications at resource-constrained wireless devices (WDs). To improve the spectrum efficiency for multi-WD computation offloading, this paper considers non-orthogonal multiple access (NOMA) assisted two-tier multi-access edge computing scenario, which exploits the computation resources of both the edge servers (ESs) and the cloudlet server (CS) deployed at different tiers. In particular, the WDs can offload partial workloads to different ESs simultaneously via NOMA, and the ESs can form a NOMA-group to further offload partially received workloads to the CS for processing. We investigate the total energy consumption minimization problem by jointly optimizing the two-tier offloading decisions, the NOMA transmission duration, and the computation resource allocation. Due to the successive interference cancellation in the NOMA and the coupling effect in two-tier offloading, the formulated optimization problem is strictly non-convex. To address this difficulty, we exploit the hierarchical relationship among the joint optimization variables, and then propose a hybrid deep reinforcement learning (HDRL) algorithm to learn two policies that determine the coupled variables, i.e., the ESs' offloading decisions and the NOMA transmission duration, respectively. Then, the remaining decision variables can be jointly optimized by using the convex optimization methods directly based on the results provided by the HDRL algorithm. Specifically, the HDRL algorithm that uses different policies to determine the coupled variables can converge faster than the existing solutions that learn a single policy to determine all variables. Experimental results are provided to validate the performance of our proposed HDRL algorithm in comparison with two other learning-based algorithms.
UR - http://www.scopus.com/inward/record.url?scp=85139260335&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85139260335&partnerID=8YFLogxK
U2 - 10.1109/WOCC55104.2022.9880599
DO - 10.1109/WOCC55104.2022.9880599
M3 - Conference contribution
AN - SCOPUS:85139260335
T3 - 2022 31st Wireless and Optical Communications Conference, WOCC 2022
SP - 138
EP - 143
BT - 2022 31st Wireless and Optical Communications Conference, WOCC 2022
PB - Institute of Electrical and Electronics Engineers Inc.
Y2 - 11 August 2022 through 12 August 2022
ER -