The publisher just completed its 22nd 5G NR benchmark study. For this endeavor they conducted a benchmark study of the Verizon 5G Ultra Wideband network (60 MHz @ Bn77) in the Twin Cities Metro area where Ericsson is the infrastructure supplier.
Highlights of the Report include the following:
Acknowledgements
This study was conducted in collaboration with Accuver Americas and Spirent Communications who provided their respective test equipment and platforms, which is identified in the report. The publisher did all the testing and analysis of the data and we are solely responsible for the commentary in the report.
Publisher Methodology
We used a recently-purchased Galaxy S21 FE smartphone with a Verizon test SIM to collect downlink/uplink performance data while driving ~500 kilometers and consuming 1.48 TB of data. We analyzed chipset diagnostic messages using time and area binning (5×5 meter grids) to provide the most meaningful results and analysis.
That was then, this is now
Although 5G Nationwide (Bn5) performance and coverage was a disappointment, Bn77 was not. Coverage greatly exceeded what we observed with Bn5 back in Oct 2020 while average throughput was 15-20x higher. Bn77 spectral efficiency was also much higher than all LTE bands and Bn5.
FDD-TDD CA and UL-256QAM is alive and well
The Verizon network supported both features which the publisher believes are critical for a successful mid-band 5G NR deployment. Greater use of FDD-TDD CA and device support for UL-256QAM are important success factors.
Bn77 vs Bn41 Coverage
Although the area covered by Bn41 was modestly larger than it was for Bn77, the Bn77 uplink coverage and performance (average throughput, spectral efficiency) was better than Bn41. Data suggests a one-for-one overlay of Bn77 on the existing LTE cell grid with no obvious detrimental impact on overall performance (e.g., performance at cell handover, etc.).
Fixed Wireless Access is in the works
Bn77 capacity in rural areas, including in areas where another broadband option doesn’t exist, greatly exceeds what is needed for eMBB. The analysis of throughput versus distance to the serving cell site also revealed rural coverage at 3.7 GHz won’t be a limiting factor. mmWave would make a great complementary solution.
Table of Contents
1.0 Executive Summary
2.0 Key Observations
3.0 Downlink Performance Results and Analysis
3.1 Performance Related Analysis
4.0 Uplink Performance and Coverage Analysis - Bn41 versus Bn77
5.0 Distance and Vehicular Speed Analysis
6.0 Test Methodology
7.0 Final Thoughts
Index of Figures & Tables
Figure 1. 5G Nationwide versus 5G Ultra Wideband
Figure 2. 5G NR NSA Drive Route
Figure 3. Technology Distribution
Figure 4. LTE Band Usage
Figure 5. Bn77 and Bn77 + LTE Average Throughput
Figure 6. Bn77 and Bn77 + LTE Throughput Distribution
Figure 7. Average Throughput by Band / Technology
Figure 8. Average Spectral Efficiency by Band / Technology
Figure 9. Distribution of MIMO Rank and Modulation Scheme19
Figure 10. Bn77 SINR Distribution
Figure 11. Bn77 RSRP Distribution
Figure 12. Bn77 RSRP versus SINR
Figure 13. Bn77 RSRP versus SINR II
Figure 14. SINR versus Modulation Usage
Figure 15. SINR versus MIMO Layer Usage
Figure 16. RSRP versus Throughput
Figure 17. SINR versus Throughput
Figure 18. Average 5G NR Throughput versus MIMO Rank 3 or Rank 4 with 256 QAM Utilization Included
Figure 19. Average 5G NR Throughput versus 256 QAM Utilization with MIMO Rank Usage Included
Figure 20. Average LTE and Bn77 RSRP
Figure 21. Distribution of LTE and Bn77 RSRP
Figure 22. LTE and Bn77 Time Series Plot
Figure 23. LTE and Bn77 Time Series Plot II
Figure 24. LTE and Bn77 Handover RSRP Values
Figure 25. 5G NR mmWave Throughput and Serving Cell PCI Time Series
Figure 26. 5G NR mmWave Throughput and Serving Cell RSRP Time Series
Figure 27. Uplink Drive Route
Figure 28. 5G NR Radio Bearer Active by Operator
Figure 29. 5G NR Throughput Greater than 1 Mbps by Operator
Figure 30. Meaningful Usage of 5G NR Uplink RBs
Figure 31. Average LTE and 5G NR Uplink Throughput - Bn41 and Bn77
Figure 32. 5G NR Uplink Spectral Efficiency - Bn41 and Bn77
Figure 33. Downlink PDSCH Traffic During Uplink Tests
Figure 34. LTE and 5G NR Median RSRP at Handover - Verizon
Figure 35. LTE and 5G NR Median RSRP at Handover - T-Mobile
Figure 36. LTE and 5G NR Tenth Percentile RSRP at Handover - Verizon
Figure 37. LTE and 5G NR Tenth Percentile RSRP at Handover - T-Mobile
Table 1. Conversion Analysis
Figure 38. Comparable RSRP Values at Handovers - Bn41 vs Bn77
Figure 39. Comparable RSRP Distribution and Average Values - Bn41 vs Bn77
Figure 40. Comparable RSRP versus Uplink Throughput and Uplink MCS - Bn41 vs Bn77
Figure 41. Comparable RSRP versus Uplink RBs and Uplink MCS - Bn41 vs Bn77
Figure 42. Comparable RSRP versus Uplink RBs and Uplink PUSCH Transmit Power - Bn41 vs Bn77
Figure 43. Distance versus Downlink Throughput - Bn77
Figure 44. Distance versus RB Normalized Downlink Throughput - Bn77
Figure 45. Distance versus RB Normalized Downlink Throughput - Bn41
Figure 46. Distance versus PUSCH Transmit Power Throughput - Bn77
Figure 47. Distance versus PDSCH RB Allocations - Bn77
Figure 48. Distance versus RSRP - Bn77
Figure 49. Speed versus RSRP and RB Normalized Throughput - Bn77
Figure 50. Average RB Normalized Throughput and RSRP by Area - Bn77
Figure 51. Actual versus Expected RB Normalized Throughput by Vehicular Speed - Bn77
Figure 52. Verizon Cell Tower
Figure 53. Verizon Cell Tower Bn77 Coverage and Total Throughput
Figure 54. T-Mobile Cell Tower
Figure 55. T-Mobile Cell Tower Bn41 Coverage and Total Throughput
Figure 56. XCAL-M in Action
Figure 57. Umetrix Data Architecture
Companies Mentioned
- Verizo
- T-Mobile
