On using deep reinforcement learning to reduce uplink latency for uRLLC services

Boutiba, Karim; Bagaa, Miloud; Ksentini, Adlen
GLOBECOM 2022, IEEE Global Communications Conference, 4-8 December 2022, Rio de Janeiro, Brazil

5G networks and beyond are shifting from dominant Downlink (DL) traffic to a more equilibrate DL/UpLink (UL) and dominant UL traffic for specific emerging services. Particularly for ultra-Reliable and Low Latency Communications (uRLLC) services, the UL latency becomes an essential factor to consider. However, current UL scheduling methods are not efficient in terms of Physical Resource Blocks (PRBs) allocation, latency,
or link adaptation. In this paper, we address the emerging challenge related to the UL latency in 5G networks and beyond. We introduce a solution based on Deep Reinforcement Learning (DRL) to dynamically allocate the future UL grant by learning
from the dynamic traffic pattern. Simulation results demonstrate the efficiency of the proposed methodology in reducing the UL latency down to 0.25 ms and ensuring the generality by reacting to different traffic models.

Rio de Janeiro
Systèmes de Communication
Eurecom Ref:
© 2022 IEEE. Personal use of this material is permitted. However, permission to reprint/republish this material for advertising or promotional purposes or for creating new collective works for resale or redistribution to servers or lists, or to reuse any copyrighted component of this work in other works must be obtained from the IEEE.

PERMALINK : https://www.eurecom.fr/publication/7003