loading page

Self-Optimized Agent for Load Balancing and Energy Efficiency: A Reinforcement Learning Framework with Hybrid Action Space
  • +2
  • Bishoy Sabry ,
  • Mariam Nabil ,
  • Ghada Alsuhli ,
  • Karim Banawan ,
  • Karim Seddik
Bishoy Sabry
Author Profile
Mariam Nabil
Author Profile
Ghada Alsuhli
Author Profile
Karim Banawan
Author Profile
Karim Seddik
Author Profile

Abstract

The paper proposes a Reinforcement Learning based agent that controls three KPIs of the mobile network to reach a maximized sum throughput of the newtork, such that the number of uncovered users is kept minimum and the energy consumed due to the MIMO technology is kept minimum as well.
The environment is a simulated mobile network using NS3.