Summary

SRG conducted a benchmark study of the Verizon Wireless 5G NR networks in Minneapolis and Chicago. Although data speeds are interesting, we focused our efforts on metrics that have far more implications on the long-term prospects for 5G NR, operating in millimeter wave frequencies and supporting eMBB use cases. This work also includes the Interdependencies between 5G NR and LTE.

Highlights of the Report include the following:

Our Thanks.  We did this study in collaboration with Accuver Americas, Rohde & Schwarz, and Spirent Communications who provided us with their respective test equipment, which we identify in the report. SRG did all the testing and analysis of the data and we are solely responsible for the commentary in the report.

Our Methodology.  We captured chipset diagnostic messages from up to two Motorola Z3 smartphones with/without the 5G Moto Mod, as well as scanner information in the two networks. Using sustained, high bandwidth data transfers (UDP and TCP), we performed walk tests and drive tests, not to mention sensitivity studies involving grip, device positioning, and sudden turns. We also did multiple user experience tests involving popular applications to determine if the applications and services took full advantage of the network capabilities.

The Results.  Although we observed sustained PDSCH Layer data speeds of 1.1 to 1.2 Gbps on numerous occasions and a peak data speed of at least 1.5 Gbps, we believe the long-term prospects for millimeter wave are better defined by parameters, such as BRSRP and BSINR, and how these parameters are impacted by distance as well as near/non-line-of-site conditions. RF reflections remain alive and well.

A bit “Wonky.”  Although the network performance was a bit wonky, it wasn’t a direct result of millimeter wave, or at least the issues we documented can easily be overcome. Device/modem stability, support for beam management, increased utilization and more efficient use of the 5G NR radio channels top the list, as does improving the interdependencies with the LTE network. The use of 1,024 element arrays in the gNB radio and/or the mixed use of wide beams and narrow beams will further improve coverage and capacity.

The Importance of LTE.  When operators deploy 5G NR FR2, they frequently use a small cell architecture. As we witnessed in our tests, and in our earlier series of small cell studies, there is substantial capacity gains with LTE and small cells. Furthermore, operators are also taking advantage of LAA. In the case of Verizon, the Motorola Z3 (w/ 5G Moto Mod) simultaneously used five LTE carriers, including three unlicensed bands in Chicago.

A Paradigm Shift. 5G NR FR2 doesn’t need to provide ubiquitous coverage, especially once operators deploy 5G NR in cellular bands and incorporate carrier aggregation. Given results of user experience tests, it is clear that 5G NR FR2 isn’t about peak speeds to individual users, but the total capacity that it offers.