RL-SF: Reinforcement Learning based Scheduling Function for Distributed TSCH Networks
Yolanda Hertita Pratama, Sang–Hwa Chung · 2022
IEEE 802.15.4e defines the Time Slotted Channel Hopping (TSCH) mode as an industrial IoT communication technology. In the event of burst traffic patterns, more cells are required to minimize packet loss, yet excessive cell allocations might result in excessive energy usage. In this study, we present a distributed scheduling technique for TSCH networks that uses reinforcement learning algorithms to calculate the required number of cells. The result demonstrates that the proposed algorithm is capable of achieving a high packet delivery ratio while preserving a reasonable network lifetime and end-to-end packet latency.