Karri Kuoppamaki of T-Mobile startled me with the answer, “My vendors tell me NR will be 25% to 50% faster than LTE.” A second source, with similar information, expects about 20%. Neither have much actual test data, and vendor salesmen are often over-optimistic. But even 20% is more than my previous reporting and I need a correction if that proves out.
Eric Xu, current Huawei Chairman, recently said consumers would find no “material difference between 4G & 5G.” That remains true. Peak speeds will be similar and total capacity only matters if the cell is near to congestion, surprisingly rare. Even then, is a fall from 300 megabits to 200 megabits meaningful to the consumer?
A 25% improvement is not a new generation.
Typical LTE installations went from 2 antennas to 4 antennas in the last 18 months. That added 50% to 75% in most cases. 256 QAM also became standard, adding ~30%. This was very important to us in the industry but no one called it a "new generation." It's the kind of modest improvement that has come along every year of two for the last two decades.
My question asked about the same generation equipment, same spectrum, and same antennas. I spoke with him afterwards and I believe we were discussing a fair comparison.
5G NR is a change in line coding, easily and inexpensively added to many new base stations. The LTE standard, I'm told, was often so close to Shannon's Limit there was little room for improvement in primary performance. (I'm not qualified to judge that.) "5G" is designed to use more spectrum and antennas, the main drivers. It allows blocks of 400 MHz in mmWave compared to 100 MHz blocks in LTE.
NR almost didn't make it into the 5G specs because telcos did not want to upgrade their LTE systems and didn't see a large advantage. Qualcomm pushed very hard because it would allow them to raise royalties. (Qualcomm royalties have recently gone up about a third, perhaps $2B/year.)
To get approval, they cut a deal with Nokia to include enough Nokia patents Nokia also could claim higher royalties. (Others are rumored.) Because Verizon intended to start in 2018 with or without a standard, the telcos rushed madly to get the standard approved and accepted the NR.
NR has only a modest effect on latency. For example, AT&T has announced 9-11 ms latency for mmWave testing. Ericsson tells me they will ship LTE with 9 ms. latency in 2018. Ericsson also announced LTE with 2 ms., not much more than the 1 ms promised with 5G. (Neither LTE nor 5G is shipping with latency that low.)
State of the art handset chips for LTE (Qualcomm) & 5G (Huawei) were both at MWC. Both companies told me the peak speed was about 2 gigabits. The performance was remarkably similar. link
A very senior engineer told me last fall NR had only a small speed advantage over LTE. However, she and others believe the new equipment, called 5G, would be much better than the older gear in the field. Duh.
5-10 times improvements are based on comparisons of older LTE equipment and future 5G equipment, obviously misleading. Often, they use more spectrum and antennas in their 5G figure than in the LTE. Nearly all the improvements in 5G (except mmWave) work in LTE as well and are being incorporated in today's equipment. Apples to apples, the differences are not large.
The most important improvement in latency is "short TTI." That was included in 3GPP Release 14 (LTE) and works for both LTE & 5G. There are some small differences in throughput in some cases. Guard bands may be larger in many LTE deployments and some other changes may have an impact.
This article is about whether the NR (New Radio) line code makes a difference in LTE frequencies. The slower low band and mid-band will be 80% to 90% of "5G" for the next few years. Millimeter Wave is not supported in LTE and is a major improvement. 5G mmWave in 2018 testing is delivering two or three times the total capacity of 2018 LTE and somewhat more speed to the user. That will probably improve.
The hype has been ridiculous. It's easy to understand why vendors are pushing, but the regulator response is shameful. At the FCC, Jessica Rosenworcel, Ajit Pai, and Mike O'Rielly continue to make unsupportable claims. I know them all to be very intelligent and I believe them honorable. It's a frightening example of "confirmation bias." I hear similar from Ansip Andrus & Roberto Viola at the EU and also from MIIT.
They are making crucial decisions in order to "promote 5G." They have impossible expectations for what 5G will deliver, egged on by telcos wanting concessions worth 10's of billions or more.