The publisher just completed its 21st 5G NR benchmark study. For this study, the T-Mobile commercial 5G NR network in the D.C.-Maryland suburbs was leveraged. The large cell cluster where testing was done supported a wide range of different technologies, including support for Non-Standalone (NSA) and Standalone (SA) network architectures - Band n41 and/or Band n71. More interestingly, many of the cell sites supported 5G NR carrier aggregation (CA) involving two 5G NR carriers in Band n41 (TDD-TDD CA) or they paired a single radio channel in Band n71 with a single radio channel in Band n41 (FDD-TDD CA). A few mobile operators have already launched TDD-TDD CA in other parts of the world but to the best of the publisher's knowledge T-Mobile is the first operator in the world to deploy FDD-TDD CA functionality in a commercial network. Ericsson is the vendor in the network tested.
With TDD-TDD CA (100 MHz + 20 MHz) and NSA, the publisher observed a peak physical layer (PDSCH) throughput exceeding 1.7 Gbps with 85% of the throughput due to 5G NR. There were two LTE carriers contributing just over 260 Mbps when this measurement was made.
When the phone operated in FDD-TDD CA mode, the total throughput was lower - in part due to less total bandwidth in the 5G NR FDD channel (2x10 MHz) and in part to how T-Mobile has optimized its network. However, the true benefit of FDD-TDD CA has very little to do with total throughput and everything to do with Bn41 coverage extension not to mention making the transition to a complete 5G NR SA network architecture possible.
Although the publisher dedicates a chapter to TDD-TDD 5G NR CA performance in this report, the focus is on 5G NR FDD-TDD CA. Specifically, the publisher shows how the use of Band n71 (600 MHz) as the primary cell (P Cell) results in better performance of 5G NR in Band n41 (2.5 GHz), including higher data speeds close to the cell site, and, most importantly, higher data speeds and extended Band n41 coverage at the edge of the cell. Given the Gigabit-persecond data speeds observed, they also looked at the impact of highspeed data transfers on the smartphone’s battery temperature. The publisher document the increase in battery temperature as a function of the amount of downloaded data/transfer time with ideal 5G NR and LTE radio conditions. In the absence of exogenous factors, a 5G NR smartphone can download a tremendous amount of data before reaching the “danger zone,” but exogenous factors can’t be ignored. Finally, they include an analysis of cell handover times involving 5G NR NSA and 5G NR SA modes of operation, as well as real-world latency and jitter measurements that go beyond a simple ping to an operator’s data center located within its network. This analysis demonstrates the dramatic reduction in handover times [and fewer handovers] in the absence of LTE, as well slight differences in latency and jitter, although still well short of what is possible with uRLLC and Mobile Edge Compute.
Highlights of the Report include the following:
- CA Results. The publisher observed the same incremental benefits with 5G NR CA that exist with LTE CA. In the case of T-Mobile, the operator is using 100 MHz for the Primary Cell (P Cell) and 20 MHz for the Secondary Cell (S Cell), both in Band n41 (2.5 GHz). Peak speed was >1.7 Gbps with LTE (NSA), but the benefit of using TDD-TDD SA is that 5G NR performance is better than with TDD-TDD NSA when there is loading on the LTE network. Further SA allows the operator to move away from LTE and its limitations.
- FDD-TDD CA is more Interesting. With FDD-TDD CA, Band n71 is used for the P Cell with Band n41 (100 MHz channel) serving as the S Cell. With this configuration, greater Band n41 coverage was documented as well as instances of higher throughput closer to the cell.
- Smartphone Battery Temperature. The publisher analyzed the impact of sustained downlink data transfers on the smartphone battery temperature. Ignoring exogenous factors, it is possible to download significant amounts of data before the smartphone does a 5G NR radio link failure (RLF) due to temperature, but exogenous factors cannot be ignored.
- Handovers, Latency and Jitter. Significant improvements in handover times were documented as well as some differences in latency and jitter with the NSA mode than with the SA mode.
1.0 Executive Summary
2.0 Key Observations
3.0 Downlink Performance Results and Analysis
4.0 FDD-TDD Coverage Extension Analysis
5.0 Thermal and Handover Analysis
5.1 Thermal Analysis
5.2 Handover Analysis
5.3 Latency and Jitter
6.0 Test Methodology
7.0 Final Thoughts
Index of Figures & Tables
Figure 1. Testing Routes in and around Washington, D.C.
Figure 2. 5G NR NSA Drive Route
Figure 3. 5G NR SA Drive Route
Figure 4. 5G NR Standalone Throughput Distribution and Average Values
Figure 5. 5G NR Non-Standalone Throughput Distribution and Average Values
Figure 6. 5G NR Standalone PDSCH Throughput Versus SINR
Figure 7. 5G NR Non-Standalone PDSCH Throughput Versus SINR
Figure 8. 5G NR Standalone PDSCH Throughput Versus RSRP
Figure 9. 5G NR Non-Standalone PDSCH Throughput Versus RSRP
Figure 10. 5G NR Non-Standalone P Cell and S Cell PDSCH Throughput Versus RSRP
Figure 11. 5G NR Non-Standalone P Cell and S Cell PDSCH Throughput Versus SINR
Figure 12. 5G NR Non-Standalone PUSCH Transmit Power Versus RSRP
Figure 13. FDD-TDD Versus TDD Only Band n41 Throughput Versus RSRP
Figure 14. FDD-TDD Versus TDD Only Band n41 Throughput Versus RSRP - enhanced
Figure 15. FDD-TDD Versus TDD Only Band n41 Downlink MCS Versus RSRP
Figure 16. FDD-TDD Versus TDD Only P Cell PUSCH Transmit Power Versus Band n41 RSRP
Figure 17. FDD-TDD Versus TDD Only P Cell PUSCH Transmit Power Versus RSRP
Figure 18. FDD-TDD Versus TDD Only RSRP Versus Distance
Figure 19. TDD-TDD Standalone P Cell RSRP Versus Distance Scatter Plot
Figure 20. FDD-TDD P Cell RSRP Versus Distance Scatter Plot
Figure 21. TDD-TDD Non-Standalone P Cell RSRP Versus Distance Scatter Plot
Figure 22. FDD-TDD Versus TDD Only Transmit Power Versus Distance
Figure 23. 5G NR TDD-TDD Bn41 Standalone PDSCH RB Normalized Throughput Versus Distance Scatter Plot
Figure 24. 5G NR TDD-TDD Bn41 Non-Standalone PDSCH RB Normalized Throughput Versus Distance Scatter Plot
Figure 25. Downloaded 5G NR + LTE Data, and Battery Temperature versus Time
Figure 26. Downloaded 5G NR + LTE Data, and Battery Temperature versus Time
Figure 27. Downloaded 5G NR + LTE Data, and Battery Temperature versus Time
Figure 28. Sample Outcomes
Figure 29. NSA and SA Handover Times
Figure 30. NSA and SA Handover Times
Figure 31. NSA and SA Downlink One-Way Latency and Jitter Times
Figure 32. NSA and SA Uplink One-Way Latency and Jitter Times
Figure 33. XCAL-M in Action
Figure 34. Umetrix Data Architecture