TY - GEN
T1 - Inerrant-Bits Consumption for Over-the-Air Computation-based Multi-tier Computing
AU - Wan, Junjie
AU - Zhang, Fan
AU - Wang, Kunlung
N1 - Publisher Copyright:
© 2023 IEEE.
PY - 2023
Y1 - 2023
N2 - In this paper, we propose an novel performance metric inerrant-bits consumption (IBC) in an Over-the-Air computation (AirComp) based multi-tier computing system to evaluate the computation efficiency, which is based on the total task computation and communication energy consumptions. Then we solve an optimal transmit power design problem to minimize the IBC of an AirComp system under the peak power constraint of sensors. To address the non-convexity of the optimization problem, an efficient optimization algorithm is developed to obtain a closed-form solution. Given the power allocation result, we obtain the task offloading result. By comparing our optimal power transmission policy with two benchmark policies in simulation, we show that our proposed scheme outperforms the baseline schemes with different SNR and the sensor number setups.
AB - In this paper, we propose an novel performance metric inerrant-bits consumption (IBC) in an Over-the-Air computation (AirComp) based multi-tier computing system to evaluate the computation efficiency, which is based on the total task computation and communication energy consumptions. Then we solve an optimal transmit power design problem to minimize the IBC of an AirComp system under the peak power constraint of sensors. To address the non-convexity of the optimization problem, an efficient optimization algorithm is developed to obtain a closed-form solution. Given the power allocation result, we obtain the task offloading result. By comparing our optimal power transmission policy with two benchmark policies in simulation, we show that our proposed scheme outperforms the baseline schemes with different SNR and the sensor number setups.
KW - Over-the-air computation
KW - computation efficiency
KW - multi-tier computing
KW - power allocation
UR - https://www.scopus.com/pages/publications/85163361143
U2 - 10.1109/CISCE58541.2023.10142476
DO - 10.1109/CISCE58541.2023.10142476
M3 - 会议稿件
AN - SCOPUS:85163361143
T3 - 2023 5th International Conference on Communications, Information System and Computer Engineering, CISCE 2023
SP - 346
EP - 350
BT - 2023 5th International Conference on Communications, Information System and Computer Engineering, CISCE 2023
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 5th International Conference on Communications, Information System and Computer Engineering, CISCE 2023
Y2 - 14 April 2023 through 16 April 2023
ER -