@article{MCE747361,
title = "Deep Reinforcement Learning-Based Content Caching for Private 5G Network",
journal = "The Journal of Korean Institute of Communications and Information Sciences",
year = "2023",
issn = "1226-4717",
doi = "10.7840/kics.2023.48.3.327",
author = "JoonyoungLim,DongjuKim,YounghwanYoo",
keywords = "Local 5G, Private 5G, Deep reinforcement learning, Network cache",
abstract = "Although the demands for local 5G network has increase along with the 4th industrial revolution, current network operation techniques and systems cannot efficiently manage local 5G networks, and suitable system for those networks are necessary. Therefore, we propose a deep reinforcement learning based caching system to reduce backhaul overload and increase user QoS in local 5G for efficient network resource utilization in eMBB targeted local 5G networks. The proposed system considers replacement policies in the stage of cache allocation, and its performance is compared with existing caching strategies combined with cache allocation algorithm and cache replacement policy. The simulation result shows the proposed system has 20% higher performance in both cache hit ratio and average network latency than conventional systems."
}