Deep Reinforcement Learning Model Improves Handover Success and Reliability in Simulated 4G LTE Networks

Written by

in

In a 4G LTE simulated network calibrated with data from 910 base stations in Nigeria’s Niger Delta, a deep reinforcement learning model increased average handover success to 97.3%, 15.5 percentage points higher than the 3GPP fixed-parameter baseline. Using the Multi-KPI Reward Function to integrate five metrics, the model reduced the simulated dropped call rate to 1.8% and ping-pong handovers by 34.7%. These results are based solely on simulation evaluations.

Source:BIOENGINEER.ORG · Read original

Comments

Leave a Reply

Your email address will not be published. Required fields are marked *