Talk:5G NR frequency bands

Band n89

Band n89 is only 25 MHz wide yet the 3GPP is allowing an n89 channel to be deployed up to 50 MHz wide in their latest release of the 38.101 spec which to me seems like a typo. Should we leave it as is or remove the 50 MHz bandwidth from the article Joshua Shah (talk) 01:13, 19 January 2022 (UTC)

There probably is something fishy about n89 in the TS. Neither n5 nor n26 uses SCS of 60 kHz, so it's strange that n89 does to achieve a bandwidth beyond its channel width. ebahapo (talk) 04:30, 20 January 2022 (UTC)

So should we leave it or wait until ~ March for 3GPP rel 17.5 to be released? Joshua Shah (talk) 23:52, 22 January 2022 (UTC)

There might be a revision before that, but it's hard to go against the authoritative source for this information. As a temporary solution, I added sic to it. What do y'all think? ebahapo (talk) 02:52, 23 January 2022 (UTC)

Yeah I think that'll be fine. Joshua Shah (talk) 00:40, 30 January 2022 (UTC)

Surely a reasonable approach. Nightwalker-87 (talk) 19:35, 31 January 2022 (UTC)

Bands n46, n96, and n102 Change Note from LAA to NR-U

In the table Frequency Range 1, bands n46, n96, and n102 have notes of "LAA" but, LAA is a 4G LTE capability. The correct 5G term is NR-U or 5G NR-U. I proposed to change "LAA" to "NR-U" in those entries. NR-U is mentioned at 5G#Deployment but may need a more comparable page to LAA. Reference IEEE document 9277664 "5G New Radio Unlicensed: Challenges and Evaluation" Mrdvt92 (talk) 12:28, 26 May 2022 (UTC)

Support: Yes, I agree on that. Nightwalker-87 (talk) 16:19, 26 May 2022 (UTC)

Band Names

Do we have a source for the band names listed here? For example, "SMH" for 700MHz, or "C-Band" for n77? I looked at the 3GPP's release TS 38.101, but they don't actually list names for any of these bands, so where did the names in the article come from? They need to be sourced, otherwise they'll be removed. 73.128.151.200 (talk) 22:42, 22 June 2022 (UTC)

I agree, there is no source for common names in 3GPP. Where did this information come from? Sheytoon123 (talk) 12:57, 25 June 2022 (UTC)

Common industry names often used by regulatory bodies and industry publications. ebahapo (talk) 16:43, 25 June 2022 (UTC)

If you follow the linked pages from some terms, like C band, you'll find their source. ebahapo (talk) 16:44, 25 June 2022 (UTC)

"Common industry names often used by regulatory bodies and industry publications." -- Okay... but you still need a source. You can't just make up your own names without citing a source for them. I've never heard anyone refer to 700MHz as "SMH", and "C-Band" is a term limited only to the US. Governments outside the US don't refer to the 3.45-4.0GHz bands as "C-Band". The Canadian government simply refers to them as "3500MHz" and "3800MHz", which is also common across Europe. Technically, 3.45GHz isn't C-Band, which is defined as 4-8GHz by IEEE. 73.128.151.200 (talk) 20:04, 25 June 2022 (UTC)

C-Band is not defined as 3300–4200, 3300-3800, 4400–5000 in any industry publication or regulatory body. If you have a reference, please provide it.

For the most part, it's a US term, and even vendors like Ericsson and Samsung have publicly accessible pages stating 3.7-4.2 GHz. See examples:

https://www.ericsson.com/en/press-releases/6/2020/ericsson-demonstrates-industry-first-live-c-band-network-with-54-gbps-throughput

https://www.samsung.com/global/business/networks/insights/press-release/0419-samsung-introduces-complete-c-band-network-solutions-portfolio/

This causes confusion with the public, who believe n78 deployment in countries like Canada are C-Band.

We could add a note to say C-Band is only for certain countries (like US), or remove the column entirely, or remove references to C-Band for n77, n78, n79. Sheytoon123 (talk) 16:03, 26 June 2022 (UTC)

I don't think the entire column needs to be removed, but each "common name" needs to be individually sourced. If there's no source, the name should be changed or removed. I would just go by whatever the government agency who auctions the spectrum refers to it as. For example, the terms "AWS" and "PCS" are used in North America, and "Cellular" for 850MHz. It's less clear with bands like n41. The FCC in the US calls this band BRS/EBS. Canada calls this BRS. Other countries in Asia typically just refer to it as "2500MHz" or "2600MHz". I'll try to find as many sources as I can and source what I can find. 73.128.151.200 (talk) 18:41, 26 June 2022 (UTC)

That sounds good to me. Sheytoon123 (talk) 12:18, 27 June 2022 (UTC)

Do you think it's a bit early to label 3.45-3.55 GHz range for US as "3.45 DoD"? The 3GPP working group notes don't specifically call it "3.45 DoD". Would it better to remove this until there is more alignment? Adding "3.45 DoD" may lead to confusion for people when they think of n78. Sheytoon123 (talk) 00:01, 28 June 2022 (UTC)

There is a 3GPP meeting document which refers to "n77-DoD" several times. It's the Word document I linked in the sources. The spectrum in the US came from the Department of Defense. Officially, the FCC only calls it "3.45 GHz", but it's not the same block of spectrum as C-Band, so it would be inaccurate to call it C-Band. Carriers in the US are using band n77 for both 3.45-3.55 and 3.7-3.98. I know the networks in Canada decided to use n78 for their 3.45 spectrum, but the government of Canada only refers to it as "3500 MHz". 73.128.151.200 (talk) 04:04, 28 June 2022 (UTC)

Are you referring to this document: https://www.3gpp.org/ftp/tsg_ran/WG4_Radio/TSGR4_99-e/Inbox/Drafts/%5B99-e%5D%5B161%5D%20US_n77/Round%201/Summary_161_1st%20round_v17_Sams_Google.docx

Honestly I don't see "n77-DoD" defined in there. This document is essentially meeting minutes from one particular meeting between some vendors. I see a single comment from MediaTek that refers to "US-n77 DoD".

In my mind, the purpose of the "common name" column would be to highlight names that are more common than this. Just my 2 cents. Sheytoon123 (talk) 23:16, 28 June 2022 (UTC)

There are numerous references to "DoD" spectrum in that document. Do a Ctrl+F for "DoD". What other names are more common? 3.45GHz isn't C-Band, so it would be incorrect to label the entire band as C-Band. The US carriers use n77 for both 3.45 and 3.7GHz. 73.128.151.200 (talk) 15:38, 30 June 2022 (UTC)

I would remove references to DoD entirely, and just add "C-Band (US)" for n77 and n78 Sheytoon123 (talk) 22:31, 30 June 2022 (UTC)

That wouldn't be accurate, though. Band n78 isn't used in the US, and isn't C-Band. Only n77 is used in the US. It's often called C-Band, because it extends to 4.2GHz which is (barely) into C-Band's range of 4-8GHz. 73.128.151.200 (talk) 00:03, 2 July 2022 (UTC)

The SMH naming derives from common literature, e.g.: https://books.google.de/books?id=HRQmBgAAQBAJ&pg=PA437&lpg=PA437&dq=upper+SMH+band&source=bl&ots=JeBPNzfQS6&sig=ACfU3U2dcLgFLzUX5vytVdwW2gH2Egixwg&hl=de&sa=X&ved=2ahUKEwjOlZrP79r4AhWt57sIHcKCDQIQ6AF6BAg1EAM#v=onepage&q=upper%20SMH%20band&f=false As it is an established reference to the respective bands, there appears to be no need for individual references. Nightwalker-87 (talk) 18:58, 2 July 2022 (UTC)

@ 73.128.151.200: fair point. I'm ok with putting "C-Band (US)" for n77.

@Nightwalker-87 You've made a lot of changes without providing explanations.

1) Why did you label n90 as a subset of n41, and you also removed n90 as a superset of n38?

2) Why are you removing the references that 73.128.151.200 provided?

3) Why do you think n79 should be called C-Band when Ericsson and Samsung are saying it doesn't fall in that range? Why do you think C-Band is a global term when it's mainly used in the US for n77? Sheytoon123 (talk) 00:46, 4 July 2022 (UTC)

It is precisely clear to me that 73.128.151.200 and Sheytoon123 are the same person and are sockpuppets of User:Dnywlsh. Joshua Shah (talk) 15:26, 5 July 2022 (UTC)

Well possible - at least it appears to be one of those type of conversations we've seen before (questionable, reproachful & with a vague level of argumentation), what may likely underline your statement. Nightwalker-87 (talk) 18:36, 6 July 2022 (UTC)

You can't revert other people's edits without discussion here. The vast majority of these "common names" are not sourced. No, no one refers to 700MHz as "SMH". You need to provide sources for these names, otherwise the entire column may be removed. I'm getting the site admins involved at this point. 73.128.151.200 (talk) 22:33, 7 July 2022 (UTC)

Just because 2 people on the internet disagree with arbitrary naming and want to see references, it doesn't mean we are the same person. I'm in Canada, 73.128.151.200 appears to be American. Not that it should matter. Sheytoon123 (talk) 15:11, 9 July 2022 (UTC)

Can we just stick to providing references please? I still haven't seen answers to my questions above. Sheytoon123 (talk) 18:15, 8 July 2022 (UTC)

@Sheytoon123 See the provided link to printed literature I provided above. To my knowledge several similar references exist in other books.

... or here: https://www.5gradio.com/5g-technology/5g-nr-frequency-bands/

... or here: https://www.rfwel.com/us/index.php/5g-nr-frequency-bands

... or here: https://www.hocell.com/newsinfo/479211.html?templateId=1133604

Nightwalker-87 (talk) 19:44, 8 July 2022 (UTC)

@Nightwalker-87 Those are blogs, the first one and third one essentially copied the Wikipedia tables, which makes them circular references. I don't think that is valid. The second one doesn't list common names for n77, n78, n79.

I've responded to question 1 in the dedicated discussion for n90 separately.

Mind if I ask you what your technical experience is with telecom and RAN? Sheytoon123 (talk) 15:08, 9 July 2022 (UTC)

As mentioned and cited before as an example (18:58, 2 July 2022), there are also references existing in printed literature. However, as they as such do stand for themselves without any further need for explanation and would be in accordance with WP:CS, I'll will not participate any further in this discussion which appears not to focus on WP:DR based on WP:RS inevitably leading to WP:DISCFAIL and from my perspective also violating WP:AGF as well. Nightwalker-87 (talk) 15:28, 9 July 2022 (UTC)

Your citation from July 2 does not mention C-Band. I've provided 2 references from Ericsson and Samsung, who are actual RAN vendors in US with real world deployments, and they have defined C-Band as 3.7-4.2 GHz. This is outside the range of n79.

Labeling n79 as C-Band is inappropriate for cellular deployments. Sheytoon123 (talk) 19:02, 9 July 2022 (UTC)

From US FCC: https://docs.fcc.gov/public/attachments/DOC-374358A1.pdf

"Specifically, the Wireless Telecommunications Bureau announced the grant of 5,676 licenses in the 3.7 GHz service (3.7 to 3.98 GHz, also referred to as the C-band) following completion of Auction 107 earlier this year." Sheytoon123 (talk) 01:54, 11 July 2022 (UTC)

Additional references from FCC: https://docs.fcc.gov/public/attachments/FCC-20-22A1_Rcd.pdf

"Today, we expand on these efforts to close the digital divide and promote U.S. leadership in the next generation of wireless services, including 5G wireless and other advanced spectrum-based services, by reforming the use of the 3.7-4.2 GHz band, also known as the C-Band. By repacking existing satellite operations into the upper 200 megahertz of the band (and reserving a 20 megahertz guard band), we make a significant amount of spectrum—280 megahertz or more than half of the band—available for flexible use throughout the contiguous United States, and we do so in a manner that ensures the continuous and uninterrupted delivery of services currently offered in the band."

"We adopt rules to add a primary non-Federal mobile, except aeronautical mobile, allocation to the 3.7-4.0 GHz band nationwide."

"Next, we designate 280 megahertz of C-band spectrum (3.7-3.98 GHz) throughout the contiguous United States to be cleared for auction plus another 20 megahertz (3.98-4.0 GHz) to be cleared to serve as a guard band."

"We next adopt rules to limit FSS operations to the 4.0-4.2 GHz band in the contiguous United States. To accomplish this goal and make the 3.7-4.0 GHz band available for terrestrial wireless use, we use our authority under section 316 of the Communications Act to modify the existing FSS licenses and market access authorizations held by space station operators in the band."

To summarize these points, FCC has defined C-Band as 3.7-4.2 GHz (aligned with Ericsson and Samsung), while further breaking that down as follows:

1) 3.7-3.98 GHz to be used for mobile services, which was auctioned by FCC in 2021

2) 3.98-4 GHz to be used as guard band between mobile and satellite services

3) 4-4.2 GHz to be used by fixed satellite service (FSS) providers Sheytoon123 (talk) 02:11, 11 July 2022 (UTC)

Additional references from Innovation, Science and Economic Development Canada (ISED):

https://www.ic.gc.ca/eic/site/smt-gst.nsf/eng/sf11437.html

"180. In Canada, the 3700-4200 MHz band is licensed for use by fixed satellite service for the delivery of telephony and Internet in northern and remote communities. The band is the downlink portion of C-band fixed satellite systems, paired with 5925-6425 MHz as the uplink."

https://www.ic.gc.ca/eic/site/smt-gst.nsf/eng/sf11627.html

"41. The 3700-4200 MHz band (space-to-Earth) is paired with the 5925-6425 MHz band (Earth-to-space) and together are called the “C-band.”"

In the context of mobile networks, 3700-4200 MHz is the frequency range of interest that falls within C-Band definition. Sheytoon123 (talk) 02:24, 11 July 2022 (UTC)

The sources that you have provided have nothing to do with n79 as they deal with the frequencies between 3.7 - 3.98 GHz which the U.S. defines as C-Band. The IEEE defines C-Band as frequencies between 4 - 8 GHz (which n79 falls within) but it also shows in C_band_(IEEE)#Differences in frequency_range by geographic area that 3.4 GHz was the lowest frequency that was used by satellites that were considered to be using C-Band. You cannot just come here and decide for everyone that "In the context of mobile networks, 3700-4200 MHz is the frequency range of interest that falls within C-Band definition," but need a reliable source that explicitly says that. Nightwalker-87 included an article from the Android Authority that states, "More specifically, C-Band refers to radio frequency bands between 4 and 8GHz, according to the Institute of Electrical and Electronics Engineers (IEEE) — although in the case of the US, the FCC designates the range as between 3.7 and 4.2GHz. Technically, that’s already classed as a sub-6GHz spectrum, and the International Telecommunication Union (ITU) has divided up C-band into the n77 (3.7GHz), n78 (3.5GHz), and n79 (4.7GHz) bands you’ll find in use around the world right now. So you could already be reading this on a C-Band 5G connection." This article is not specific to what the U.S. defines as C-Band only and therefore we have to include the viewpoint of other countries and International bodies like the ITU. Joshua Shah (talk) 17:00, 20 July 2022 (UTC)

Actually the sources I provided mention the following, word for word:

Ericsson: "C Band is the much-anticipated mid-band spectrum, consisting of 500 megahertz between 3.7-4.2 GHz"

Samsung: "C-Band refers to mid-band spectrum ranging from 3.7GHz to 4.2GHz"

FCC: "3.7 to 3.98 GHz, also referred to as the C-band"

FCC: "the 3.7-4.2 GHz band, also known as the C-Band"

FCC: "Next, we designate 280 megahertz of C-band spectrum (3.7-3.98 GHz) throughout the contiguous United States to be cleared for auction plus another 20 megahertz (3.98-4.0 GHz) to be cleared to serve as a guard band."

Why can't my references stay there? How come you are allowed to decide for everyone which references can stay?

Despite its name, Android Authority is not an authority in this area. Ericsson, Samsung and FCC are.

C-Band on mobile networks does not refer to 4-8 GHz. If it did, LTE LAA should be classified as C-Band. Are you suggesting n46, n47, etc. should also be called C-Band?

Can you share where ITU has mentioned C-Band? Sheytoon123 (talk) 21:36, 20 July 2022 (UTC)

Technically, LAA between 5 GHz is C Band as it falls between 4 - 8 GHz so you won't be wrong but those specific frequencies were given a name specifically UNII. C-Band is C-Band no matter what form of technology you deploy the spectrum on. Ericsson and Samsung are referring to the U.S. deployments in those articles that you referenced and are marketing their products for that country. You cannot forget that Ericsson, Nokia and Samsung also supply C Band equipment between 3.4 - 3.8 GHz for Europe/Asia. I removed your references because they have nothing to do with n79 specifically. Joshua Shah (talk) 23:00, 20 July 2022 (UTC)

I disagree with that. My references clearly defined C-Band as deployed in cellular networks, which is what this Wikipedia page is all about (i.e. 5G). n79 is not C-Band. No mobile operator is calling n79 C-Band, while US operators are calling n77 C-Band. My suggestion is to add my references back, because I believe they are very relevant and valid. I also think we should label n77 as "C-Band (US)", and remove common names for n78 and n79. Is there some requirement that every single 3GPP band must have a common name on Wikipedia?

You still haven't showed where ITU mentions C-Band. Sheytoon123 (talk) 00:11, 21 July 2022 (UTC)

The ITU has released it in one of their frameworks for 5G. The articles from PCMag, Android Authority etc. is enough for a wikipedia article but if you really have the time to hunt down for that document to please your mind please feel free to do so. N77 is also used in China and Japan. You completely missed the point that I was trying to explain above. Joshua Shah (talk) 01:36, 22 July 2022 (UTC)

ITU is actually a fairly reputable organization, that's why I was asking. Since you haven't provided a source for it, I won't take that comment seriously.

Blogs like PC Mag, Android Authority, etc are not reliable sources for telecom. Likewise, IEEE has no relevance in cellular networks. The reliable sources for the telecom industry are 3GPP, RAN vendors, telcos/operators, UE/chipset manufacturers, and national regulators.

C-Band is not a common name for n79 by any stretch of the imagination.

C-Band is not a 5G term that is used widely (if at all) outside of the US. In terms of 5G bands, which this Wiki page is about, C-Band is mainly used by the US to refer to n77. Sheytoon123 (talk) 01:57, 22 July 2022 (UTC)

I've written extensively on RAN topics on the internet, it's an area of expertise for me and one that I am happy to contribute to. Unfortunately it seems my expertise is not valued here. And for some reason, some people have accused me of being a sock puppet. I'll continue with my contributions on HowardForums, RFD, PM community, MobileSyrup and Reddit, but I hope you guys will reconsider this closed view towards outside opinions. Otherwise the quality of information will suffer. Sheytoon123 (talk) 02:04, 22 July 2022 (UTC)

I'd like to invite you to read WP:RS, especially WP:RSPRIMARY Joshua Shah (talk) 02:17, 22 July 2022 (UTC)

Will do. And please don't take this the wrong way, but if you or anyone here is interested in reading about how 4G and 5G technologies work, I have lots of posts on other websites that have been generally well received. Let me know! Sheytoon123 (talk) 16:29, 22 July 2022 (UTC)

Bands n38, n41, n90

Right now it's showing n90 as a subset of n41, but that's incorrect. They cover the same frequency range, but the channel raster is more flexible on n90. I would propose to list n38 as a subset of n41 and n90. What does everyone think? — Preceding unsigned comment added by Sheytoon123 (talk • contribs) 13:05, 25 June 2022 (UTC)

n38 supports the 5 MHz bandwidth that's not found in n41. n41 supports more bandwidths than n90 though. ebahapo (talk) 16:39, 25 June 2022 (UTC)

I understand the channel bandwidth difference, but that wasn't what I was referring to. n90 is not a subset of n41. Sheytoon123 (talk) 16:05, 26 June 2022 (UTC)

From my point of view the technical characteristics of two bands in the same frequency range should be compatible in order to define if a band is a subset or not. Nightwalker-87 (talk) 19:07, 2 July 2022 (UTC)

What do you mean by compatible? n41 and n90 are the exact same frequency range with different rasters. Sheytoon123 (talk) 03:04, 9 July 2022 (UTC)

From TS 38.101 Table 5.2-1 NOTE 5:

Unless otherwise stated, the applicability of requirements for Band n90 is in accordance with that for Band n41; a UE supporting Band n90 shall meet the requirements for Band n41. A UE supporting Band n90 shall also support band n41.

From TS 38.101 Table 5.4.2.3-1:

n38 ΔFRaster = 100 kHz

n41 ΔFRaster = 15, 30 kHz

n90 ΔFRaster = 15, 30, 100 kHz

n90 is not a subset of n41. n90 was specifically created to harmonize the raster granularity. For example, operators who own n7 (FDD) and n38 (TDD) were not able to deploy their n38 spectrum as n41 in the centre of their spectrum holdings due to n41 raster limitation. They also couldn't deploy it as n38 due to lack of UE ecosystem. n90 solves this problem. Sheytoon123 (talk) 15:02, 9 July 2022 (UTC)