Interference Assessment for the Spectrum Sharing Between IMT-2020 and Inter-satellite Service

Bo Li, Zhaojun Qian, Shuaijun Liu, Chaowei Wang, Yinghai Zhang, Weidong Wang

TL;DR

This paper investigates whether 5G (IMT-2020) networks can safely share the 25.25-27.5 GHz millimeter-wave band with existing Data Relay Satellite (DRS) systems. Through mathematical modeling, the authors find a stark asymmetry: while satellite-to-ground interference is negligible for 5G users, the aggregate interference from terrestrial 5G base stations severely exceeds the permissible limits for geostationary satellites, making direct spectrum sharing highly problematic.

What Problem Does This Paper Address?

As 5G (IMT-2020) networks roll out, they require massive amounts of contiguous bandwidth to support high data rates, pushing cellular communications into higher millimeter-wave (mmWave) frequencies. The World Radiocommunication Conference (WRC-15) identified the 25.25-27.5 GHz band as a prime candidate for 5G expansion.

However, this spectrum is not empty. It is already allocated on a co-primary basis to Inter-Satellite Services—specifically, for the "return inter-orbit link" where Low Earth Orbit (LEO) spacecraft transmit data up to geostationary Data Relay Satellites (DRS). The core problem this paper tackles is coexistence: If we deploy thousands of 5G base stations in this band, will their combined radio emissions blind the sensitive receivers on geostationary satellites? Conversely, will the transmissions from LEO satellites interfere with 5G smartphones on the ground?

Methods

The authors use a simulation-based approach to assess interference, relying on standard mathematical models rather than physical field tests.

  • System Modeling: They model the 5G network using 3GPP standards, specifically simulating a macro-cell network with tri-sector layouts. Crucially, they account for 5G's use of massive MIMO and beamforming by modeling a 4x8 antenna array pattern.
  • Propagation Modeling: They utilize the ITU-R P.452 propagation model, which calculates line-of-sight free-space path loss, atmospheric absorption, and polarization loss.
  • Interference Scenarios: The study evaluates two specific links:
  • Ground-to-Space: Interference from 5G Base Stations (IMT-BS) to the geostationary DRS. The authors calculate "lumped interference"—the aggregate noise generated by all 5G base stations falling within the satellite's spot beam footprint.
  • Space-to-Ground: Interference from the LEO spacecraft transmitting down toward 5G User Equipment (UE).
  • Protection Criteria: They benchmark their results against the strict ITU-R SA.1155 recommendation, which states that interference to the satellite must not exceed -118.6 dBm/MHz.

Datasets

Not discussed in the paper. The study relies on standardized system parameters and mathematical propagation models provided by ITU-R and 3GPP, rather than empirical datasets.

Key Insights and Ideas

  • The Danger of "Lumped" Interference: A single 5G base station pointing its beam toward the horizon isn't a threat to a satellite 36,000 km away. However, a satellite's spot beam covers a massive geographical area. The aggregate (lumped) power of thousands of 5G base stations operating simultaneously within that footprint creates a cumulative noise floor that easily overwhelms the satellite's highly sensitive, high-gain receiving antennas.
  • Asymmetrical Coexistence: The interference is a one-way street. Because 5G handsets (UEs) have low-gain antennas and are relatively insensitive compared to satellite receivers, the signals raining down from LEO satellites do not pose a threat to terrestrial 5G performance.
  • Beamforming Isn't a Silver Bullet: Even though 5G uses directional beamforming (rather than broadcasting energy everywhere like older cellular generations), the sheer density of urban 5G deployments means enough energy still leaks upward to violate satellite protection criteria.

Key Results

  • 5G Base Station to Satellite (Failure to Coexist): The simulation showed that the aggregate interference power from 5G base stations to the DRS is always above the maximum permissible level of -118.6 dBm/MHz. This held true regardless of the elevation angle or the "Effective Area Percentage" (whether 5G covered 10% or 80% of the satellite's spot beam).
  • Satellite to 5G User (Successful Coexistence): When calculating the interference from LEO spacecraft to 5G UEs in three distinct cities (Beijing, Harbin, Wuhan), the "additional isolation required" was consistently negative (ranging from roughly -43 dB to -51 dB). A negative value means the existing spatial separation is more than sufficient; the forward feeder link can safely coexist with the 5G downlink.

Limitations and Future Work

The authors acknowledge that their simulation represents a somewhat worst-case scenario (e.g., assuming main-beam to main-beam alignment). The current model does not account for advanced, dynamic 5G network behaviors that could mitigate these issues.

For future work, they suggest investigating how advanced 5G deployment strategies, dynamic antenna beam-steering (specifically avoiding pointing beams at known geostationary orbital slots), and sophisticated power control schemes might reduce the interference footprint. These technologies will be necessary to find a viable path for the coexistence of both systems in the 25 GHz band.

Summary updated

BibTeX
@inproceedings{li2017interference,
  author = {Bo Li and Zhaojun Qian and Shuaijun Liu and Chaowei Wang and Yinghai Zhang and Weidong Wang},
  title = {{Interference Assessment for the Spectrum Sharing Between IMT-2020 and Inter-satellite Service}},
  booktitle = {SmartGIFT},
  year = {2017},
  doi = {10.1007/978-3-319-61813-5\_3},
}
Figure