Machine Learning Based Task Scheduling for Wireless Powered Mobile Edge Computing IoT Networks

In this paper, we study the joint optimization of computation mode and processing order selection in a wireless powered mobile edge computing (MEC) Internet of Things (IoT) network. In the considered network, the MEC server is integrated with a radio-frequency identification (RFID) reader that can use radio-frequency (RF) wireless power transfer (WPT) technology and each IoT device consists of sensors and a semi-active RFID tag that can harvest energy from the reader. By jointly considering computing mode selection (i.e., offloading mode or local computing mode) of each device and processing order of all devices, we aim to minimize the sum delay (i.e., the sum of energy harvesting, computational and queuing delays) for all devices under energy and delay constraints in a congestion scenario. To solve this problem, we leverage a Q-learning method to jointly optimize computing mode selection and processing order. Simulation results indicate that the proposed method can achieve up to 55.1% and 15.9% gain compared to offloading-only and local-computing-only method, respectively.