DDPG-based intelligent rechargeable fog computation offloading for IoT |
| |
Authors: | Chen Siguang Ge Xinwei Wang Qian Miao Yifeng Ruan Xiukai |
| |
Affiliation: | 1.Jiangsu Key Lab of Broadband Wireless Communication and Internet of Things, Nanjing University of Posts and Telecommunications, Nanjing, China ;2.Jiangsu Engineering Research Center of Communication and Network Technology, Nanjing University of Posts and Telecommunications, Nanjing, China ;3.Institute of Intelligent Locks, Wenzhou University, Wenzhou, China ; |
| |
Abstract: | In view of the existing computation offloading research on fog computing network scenarios, most scenarios focus on reducing energy consumption and delay and lack the joint consideration of smart device rechargeability. This paper proposes a deep deterministic policy gradient-based intelligent rechargeable fog computation offloading mechanism that is combined with simultaneous wireless information and power transfer technology. Specifically, an optimization problem that minimizes the total energy consumption for completing all tasks in a multiuser scenario is formulated, and the joint optimization of the task offloading ratio, uplink channel bandwidth, power split ratio and computing resource allocation is fully considered. Based on the above nonconvex optimization problem with a continuous action space, a communication, computation and energy harvesting co-aware intelligent computation offloading algorithm is developed. It can achieve the optimal energy consumption and delay, and similar to a double deep Q-network, an inverting gradient updating-based dual actor-critic neural network design can improve the convergence and stability of the training process. Finally, the simulation results validate that the proposed mechanism can converge quickly and can effectively reduce the energy consumption with the lowest task delay. |
| |
Keywords: | |
本文献已被 SpringerLink 等数据库收录! |
|