5G is the fifth generation of mobile networks, which promises to provide higher data rates, lower latency, and more reliable connectivity for various applications.
Satellite communications can be classified into different types based on the orbit altitude of the satellites, such as geostationary orbit (GEO), medium earth orbit (MEO), and low earth orbit (LEO). Each type has its own advantages and disadvantages in terms of latency, bandwidth, availability, and cost. Among them, GEO satellites have the highest orbit altitude (about 36,000 km), which results in high latency (about 250 ms) but also high bandwidth (up to several Gbps) and constant coverage (a single satellite can cover about one-third of the earth’s surface).
In this project, we aim to test the feasibility and performance of 5G connectivity over GEO satellites. We use a software-defined radio (SDR) platform and an open-source 5G software stack to establish the 5G radio access network (RAN) over the satellite GEO channel. We measure the key performance indicators (KPIs) of 5G over GEO, such as throughput, latency, jitter, and packet loss. We also compare the results with those of terrestrial 5G networks and analyze the impact of satellite channel characteristics on 5G performance.
The setup of the project consists of the following components:
We successfully synchronized and connected the gNB and UE over the GEO satellite channel. We achieved the MCS 17 in the downlink, which corresponds to 64-QAM modulation and rate 3/4 convolutional coding. This means that we can transmit 6 bits per symbol and have a coding rate of 0.75, which results in a high spectral efficiency and error performance.
We performed a preliminary ping test to measure the latency and packet loss of 5G over GEO. We configured the RLC layer UM mode, which is used for data transfer without error correction or re-transmission. We sent 1000 packets of 64 bytes each from the UE to the gNB and received the following statistics:
The results show that the latency is high (about 530 ms on average) due to the long propagation delay of the satellite channel (about 250 ms one way). The packet loss is low (0.2%) due to the effective FEC coding and ARQ/HARQ protocols. The pipe size is 46, which means that there are 46 packets in flight at any given time. The ipg is the inter-packet gap, which is the time interval between two consecutive packets sent by the UE. The ewma is the exponentially weighted moving average of the ipg, which reflects the variation of the ipg over time.
We also tested some TCP/IP services over 5G over GEO, such as video streaming and file transfer.
Demonstration video posted on LinkedIn.
We found that these services can operate well despite the high latency and low packet loss of the satellite channel. We still have to test the RLC layer AM mode, which is used for data transfer with error correction and re-transmission. This mode requires ACK/NACK from the other party and large data buffers, which can further increase the delay.
Further, we can use some techniques to improve the TCP performance over GEO, such as increasing the window size, using selective acknowledgments, and using fast re-transmit and fast recovery algorithms. We can also use some adaptive video streaming protocols, such as MPEG-DASH and HLS, which can adjust the video quality according to the available bandwidth and network conditions.