Joint Scheduling of eMBB and URLLC Traffic in Space-Air-Ground Integrated Networks

What is it about?

In order to maximize network performance, we use the Deep Reinforcement Learning (DRL) approach in this research to dynamically arrange flexible transmission intervals at the time slot level. By engaging with the environment, resources are dynamically and appropriately assigned to URLLC traffic. Traffic scheduling algorithms taking into account eMBB and URLLC service demands are carefully built with incentive functions. Real-time decisions are made using DRL approach to overcome uncertainty problems. The resource slicing problem is formulated as an optimization problem to enhance reliability and maximize the average data rate of eMBB consumers while respecting URLLC constraints.

Featured Image

Photo by Josh Withers on Unsplash

Photo by Josh Withers on Unsplash