Experimental Study of the Effects of RLC Modes for 5G-NTN Applications Using OpenAirInterface5G.

GLOBECOM (Workshops)(2023)

引用 0|浏览1
暂无评分
摘要
Provisioning of 5G services via Non-Terrestrial Networks (5G-NTN) has become a reality. Currently, 5G-NTN is being developed by adapting the 5G Terrestrial Network (5G-TN) protocol which presents several challenges as the satellite channel is significantly different from terrestrial channels. The retransmission mechanism is one of them. Recent works on 5G-NTN have followed the approach of disabling retransmissions from the 5G protocol stack such as Hybrid Automatic Repeat Request (HARQ) at the MAC layer and Automatic Repeat Request (ARQ) at the RLC layer while delegating it to the application layer. This approach degrades the end-user-throughput and latency. In this work, we experimentally study the effect of enabling ARQ retransmission, by using RLC Acknowledged Mode (RLC-AM) in a GEO-based 5G-NTN which is susceptible to packet loss caused by low SNR. We conduct realtime experiments to compare the effects of RLC-AM and RLC Unacknowledged Mode(RLC-Um)on applications(VoIP, video stream, file transfer) using TCP and UDP for different SNR regimes. We have used OpenAirInterface5G-NTN, which is developed to perform realtime 5G-NTN experiments, and a satellite channel emulator. We observe that for low SNR, RLC-AM performs better than RLC-UM in achieving the required bitrate and packet error rate. The reason being RLC-AM recovers the lost packets earlier in the protocol stack without delegating the retransmissions to the application layer. This becomes especially useful when HARQ is disabled. We believe that our experimental study will complement the ongoing theoretical research and help improve the procedures of the Radio Resource Control(RRC) layer in 5G-NTN-specific applications.
更多
查看译文
关键词
5G-NR,Non-Terrestrial Networks,OpenAirIn-terface
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要