Assure 5G NTN Performance Before Launch

August 1, 2023

Non-terrestrial networks (NTN) promise to finally eliminate coverage gaps across the globe. Beyond commercial applications, these fifth generation (5G) cellular networks create new use cases for critical communications and military operations. For such applications to effectively serve these mission-critical areas, however, their performance must be assured. With RF system measurement science, 5G NTN equipment developers, integrators, and network operators can reduce the time needed to create and deploy networks, using virtual engineering for first-pass success when committing to physical gear. Simulation and emulation support NTN exploration and testing, verifying current performance while supporting next-generation evolutions.

5G NTN draws many features from 5G terrestrial networks and faces many of the same challenges. Immediately apparent are the enormous data rate and reliability expectations for 5G NTN service compared to earlier satellite-communications (SatCom) networks. Delivering the required volumes of data demands the leveraging of 5G signaling fundamentals. Carrier frequencies are in the millimeter-wave range and modulation is more complex. Higher bandwidths are available, although NTN also supports lower-bandwidth applications like narrowband Internet of Things (NB-IoT).

Spectrum challenges exist as well. 5G spectrum is already tightly allocated in terrestrial networks, and researchers anticipate a move into sub-terahertz ranges for sixth generation (6G) cellular to relieve the crunch (Figure 1). An onslaught of thousands of lower-earth-orbit (LEO) satellites will add to spectrum crowding.

Another concern is the physical limits of placing high-frequency RF and computing resources in the sky. Size, weight, power, and cost (SWaP-C) become issues when moving away from immense Geosynchronous Equatorial Orbit (GEO) satellites to more compact LEO satellites.

Generally, an NTN will comprise the following four elements:

• User equipment (UE)

• gNodeB instances (base stations)

• Channels (i.e., how signals propagate across spaces and through obstacles)

• The radio access network (RAN) and core network infrastructure

5G NTNs put some or all parts of the network in constant motion. Satellite kinematics factor into connection setup, signal quality, and handovers. gNodeB instances and parts of the RAN flying aloft add to the movement of UEs at the surface. Parameters previously fixed or confined in a small range in a 5G terrestrial network suddenly become wide-ranging variables in a 5G NTN. Tracking areas, bulk delays, Doppler shifts, signal-to-noise ratio (SNR), and more elements take on dynamic characteristics.

The implications of constant motion have ripple effects. First, a UE hoping to connect to a 5G NTN must know more about the connection, and do more processing to maintain it, than in terrestrial networks. At some point, 5G UE design will likely bifurcate, with new multi-connectivity UEs designed for both a terrestrial network and NTN and legacy UEs only communicating on terrestrial networks.

Modeling and simulating channels accurately with kinematics in a high motion 5G NTN environment will be critical. The old approach of taking physical measurements and adding link budget margin will not produce reliable connections with so many variables.

Error Vector Magnitude

Error vector magnitude (EVM) is the critical quality metric for any 5G system. EVM measurements can provide a great deal of insight into the performance of digital communications transmitters, receivers, and software-defined radios. This measurement provides an overall indication of waveform distortion, representing characteristics of a device’s phase, amplitude, and noise.

The only way to properly evaluate an EVM outcome of varying parameters in any domain is to see what happens in an end-to-end interaction through all four domains: UE, gNodeB, channels, and RAN. Not every 5G NTN system supplier will develop subsystems in all four domains. When a system is under test, however, having an accurate virtual representation of it interacting with other domains is essential for first-pass success. Figure 2 shows how this emulation testbed creates a detailed visualization of EVM before and after delay pre-compensation applied in the UE. Note the uniform constellations with pre-compensation on.

Pre-Compensation for Delay and Doppler

Modeling a 5G NTN channel is not a one-time activity to verify the link budget. Bulk delays are much more prominent in 5G NTNs because of the extended distance and intervening processing between ground-based units and satellites. Doppler shift also enters the equation, with satellites moving quickly enough to affect signals. Signal quality and timing behavior vary across all operational points in a network, with direct impacts on EVM.

System Information Blocks (SIBs) defined in the Third-Generation Partnership Project (3GPP) specifications contain data needed by UEs and gNodeBs to compensate for these effects. Some of these blocks include ephemeris data for each satellite, which can be the position and velocity state vectors or the orbital parameters that will be valid at a future point in time (known as epoch time). This enables the UE to correctly estimate the satellite position over some small time window. New timers signaled in the SIB tell the UE how long it can use the information before it needs to be refreshed.

Beyond this time, the UE must assume the call to be dropped. It will need to start over with a new call and the initial cell attach process. Additional SIB information includes details about path delay and scheduling aspects. There is also information communicated about neighbor satellites to facilitate potential handovers.

The UE uses the satellite position to calculate the total path (along with timing information signaled in the SIBs) and Doppler shift. By transmitting with a pre-correction, it ensures that signals arrive at the base station with no Doppler shift. They also arrive with the correct timing relationship relative to the base station frame timing.

Digital pre-compensation is a complex interaction between UEs, satellites, and a gNodeB. A timing anomaly under specific conditions can throw off protocol decoding. Without the right tools, however, troubleshooting an error can be extremely difficult. To tackle this challenge, engineers can use simulation to model and evaluate channel behavior for a 5G NTN configuration, followed by emulation for creating prototypes of UE and gNodeB pre-compensation and other algorithms.

Early System Discovery and Exploration

Accurate multi-domain simulation of a 5G NTN link depends on four elements: an authentic representation of complex digital modulation in a 5G waveform with real-world effects, a complete model of satellite kinematics, robust modeling of RF system signal processing, and a time-correlated view of 5G protocol decoding. Figure 3 depicts a simulation approach, which combines RF system electronic design automation (EDA) and 5G system modeling platforms to deliver all of those elements.

The RF system simulation tools use the same measurement science as test and measurement instrumentation for equivalent analysis and results in virtual space. Waveforms and behavioral models in this simulation incorporate RF system knowledge, including real-world effects, with the ability to add new information from customer measurements. Using simulation brings early visibility to potential issues and the ability to explore design tradeoffs quickly, yielding valuable insights before evaluating the PHY layer in earnest.

After system-level simulation of a ground-based RF system, the next step would typically be physical prototyping. An NTN still presents a substantial risk that prototypes, once launched, would not be completely reliable and would be hard to troubleshoot. Even with lower costs of LEO satellites, any miss in performance would be expensive in terms of time and money. Emulation reduces the risk and paves the way to the first-pass success of a 5G NTN deployment. With robust channel models and pre-compensation scenarios of delay and Doppler shift from simulation, rigorous PHY-level link testing is possible with UE, channel, and gNodeB emulators.

As a 5G NTN project evolves, physical hardware can loop into its place in the end-to-end NTN virtual representation, with the same tests applied. In addition, physical measurement feedback can tune the virtual counterpart into an exact digital twin – all before NTN systems go airborne. A 5G NTN workflow emphasizing this shift into virtual engineering activities appears in Figure 4. A mature digital twin enables 5G NTN developers to explore any situation - even ones too complex or prohibitively expensive to stage in physical space. Introducing new UE designs, new protocols such as support for IoT devices, or more sophisticated regenerative satellite payloads become straightforward by inserting capability into the digital twin for evaluation first.

Defense and first responders gain another benefit from virtual engineering for 5G NTNs. Instead of exhaustive field measurements required to plan and verify a 5G NTN deployment, they can perform quick spot-check measurements. In doing so, they minimize personnel exposure to hostile conditions while speeding deployment.

The Future of NTN

Going forward, artificial intelligence (AI) may play a prominent role in vertical handovers and self-organizing inter-satellite networks. Inevitably, 6G enters the NTN conversation soon. Sub-terahertz frequency ranges and much wider bandwidths will increase the need for authentic waveforms with complex modulation in simulations. 6G also tightens EVM specifications with higher-order modulation constellations.

Across generations of specifications and technology standards, engineers working in NTN can rely on RF system measurement science in virtual space – simulation and emulation – to keep pace with this evolution. Developers embracing 5G NTN virtual-space engineering approaches will have a competitive advantage, getting off the ground faster with less risk. For military and government networks, such assurance provides a better shot at success by lowering risks to both the mission and personnel.

This article was written by Nancy Friedrich, Aerospace & Defense Industry Solutions Marketing, Keysight Technologies. For more information, visit here .

---