Verizon Verizon’s Joe Russo explains why being first with 5G SA wasn’t a priority

@TheRealDanny

THE GOAT IN 5G
Joined
Jun 17, 2007
Messages
11,123
Location
Greater Los Angeles
Carrier(s)
T-Mobile US
Device(s)
iPhone 15 Pro
Verizon has talked about moving customers to 5G Standalone (SA), but it doesn't yet have a nationwide 5G SA network. Bottom line: There’s no hurry to get to nationwide 5G SA, especially given Verizon’s vast and existing 4G LTE network.

“For me, I want to make sure that if we’re deploying a standalone core, it’s really going to benefit customers and it’s not going to be a step back in any way,” he said.

Read more: https://www.fiercewireless.com/tech/verizons-joe-russo-explains-why-being-first-5g-sa-wasnt-priority


Sent from my iPhone using HoFo
 
TBH, 5G SA only truly helps if you're completely wanting to shed any dependency on 4G LTE, and you have more of your spectrum allocated to 5G than LTE.
Eg. On T-Mobile (early deployments) 5G NSA on n71 was a good boost - especially where 5G was strong. Hanging onto a weak 5G NSA signal was detrimental. Midband 5G NSA actually was better, as it would switch to decent 4G LTE and not drag down service. Switching between 5G SA and 4G LTE has its own issues ... 5G might hang on ... too long (low band) before dropping to LTE, and you end up with a lot of flipping between 4G LTE and 5G SA.
T-Mobile's initial push for 5G SA was because they had 600MHz band almost nationwide which could be deployed and cover up many spots as '5G' on their map for ... bragging rights.
 
AKA verizon is behind and is deflecting by saying this. His interview doesn't make sense since the 4G LTE network has been getting worse and SA would alleviate some of the issues with it since more users would be off of the network.
 
This is definitely a deflection. I believe several carriers have already gone on record saying that SA has proven to be much harder than they anticipated.

The industry pressure is on, but in my view, Verizon still needs to expand n5 for SA to work well. They're already facing enough of a challenge trying to get ahead of congestion, and there are still a lot of areas with no "Nationwide 5G" or "Ultra Wideband." That's going to be an important prerequisite to SA on Verizon. I'm actually surprised that they enabled the option in the latest iOS carrier profile. Even though a support article claims it's nonfunctional at the moment, it does suggest to me that they will begin allowing those devices to use it sooner rather than later.

In my experience, a lot of Verizon macros are still B13/66 only, and n5 coverage doesn't appear very ubiquitous. Actually, somebody on Reddit with an SDR claimed a couple days ago that Verizon reverted some n5 sites in the St. Louis area to 10x10 B5, and instead deployed n2 in its place. No DSS on B5 either, just LTE. So far, it's not clear why, but it's possible that B13 congestion is so severe that they felt the need to augment it with B5 again, and replaced it with n2 where they could better afford to take the capacity hit. This is a scenario where having SA, in order to be able to use n5 as a PCC, would have likely helped.

Aside from languishing n5, in my market, a lot of the C-band antennas were either downtilted or had power reduced earlier this year. At home, I used to get weak C-band coverage and pull over 200 Mbps. But for the past few months, it doesn't cover anywhere in the neighborhood. I have to walk nearly a half mile outside to even pick up "UW" now. I'm not really sure what Verizon's long-term plan is to densify C-band, but it just feels like that was another move in the wrong direction. The most baffling thing about this shrinking Verizon C-band coverage is that AT&T has C-band on a tower right next to Verizon's, and I can pick up their C-band on the first floor.

Aside from low-band NR improvements and expansion, Verizon sorely needs SA in dense areas. I've been to quite a few places now that are outfitted with a DAS, C-band, and even mmWave, and had my phone drop to "SOS only" because of the sheer number of devices saturating the LTE PCC. Moving from 20x20 B66 (or even worse, 10x10 B13 or 5x5 B2 in my market) to a 100 MHz n77 PCC will open up a ton of capacity for more devices in these types of venues. It has been an issue for well over a year now, and I'm genuinely surprised that Verizon hasn't moved more quickly to address this particular issue. It affects thousands of people every day in stadiums, arenas, and even some festivals with enough people present.

There's no question that legacy networks, including Verizon and AT&T, will need to employ EPS Fallback for calls. Even when VoNR launches, I don't expect ubiquitous VoNR availability for at least another 3-4 years. I don't think AT&T is as concerned about launching SA as Verizon is, or should be. AT&T's LTE network is very mature and reliable at this point. With many people having devices capable of 5x or greater LTE CA, but only 1-3x NR CA; I think AT&T still prefers maximizing their non-contiguous spectrum holdings by using NSA to build CA combos that bring in 40x40 MHz or greater FDD + 120 MHz of TDD. The move to SA will mean a lot of customers who enable it might only see n5+n77 most of the time, and legacy bands will sit underutilized as more people switch to 5G devices -- specifically, 5G devices on which AT&T enables SA.

All said, standalone is still worthwhile for any operator. Aside from being a part of the transition away from the legacy LTE/Evolved Packet Core toward a pure 5GC, there's a substantial latency reduction that comes with SA that customers will realize in the form of perceived speed improvements.
 
I rather see a company get it done right the first time then rush it out to consumers. Verizon seems to think they need a little bit more time to get it done right, so be it then. I’m sure AT&T is in that situation as well.
 
I mainly just want the 5G SA feature of aggregating upper midband with mmWave, because right now 5G UW still gets wonky when you're in an area that has both n77 and n261/n260 — it would work much more smoothly if your device simply anchored to n77 and then opportunistically aggregated mmWave on top.
 
No 5g in my area and I don’t expect it for a few more years, that’s fine saves me from needing to buy another phone at some crazy high price.
 
I rather see a company get it done right the first time then rush it out to consumers. Verizon seems to think they need a little bit more time to get it done right, so be it then. I’m sure AT&T is in that situation as well.

The only thing stopping them from "doing it right" is they do not have enough dedicated 5G spectrum to make it worthwhile in non C-Band areas. Verizon just screwed themselves not being more aggressive with spectrum purchases and being quieter than they should have been about n41 with the T-Mobile/Sprint merger - Verizon totally got "screwed" the most out of that transaction as they *could* have demanded T-Mobile sell off 50% of their PCS or AWS assets in overlap markets and nabbed them up, but chose to stay mostly silent.

T-Mobile "did it right" from the start and just used 600MHz for 5G and aggressively pursued every 600MHz license they could get, and then leased the rest so they have a nationwide decent chunk of low-band 5G spectrum.

In my neighborhood, AT&T has garbage LTE performance due to super fragmented holdings but have a strong C-Band presence (somehow) - so they'd greatly benefit from SA on C-Band around here, and i'd make the carrier's Home 5G options work much, much better here, and not be stuck with sub 5Mbps upload while getting 300-500Mbps download (it's really limited by the ACK packets that can be returned at the moment)
 
In my neighborhood, AT&T has garbage LTE performance due to super fragmented holdings but have a strong C-Band presence (somehow) - so they'd greatly benefit from SA on C-Band around here, and i'd make the carrier's Home 5G options work much, much better here, and not be stuck with sub 5Mbps upload while getting 300-500Mbps download (it's really limited by the ACK packets that can be returned at the moment)
Very similar situation here. I'm sure you know AT&T has ended up with more fragmented spectrum holdings than any other US carrier due to the way they amassed spectrum from so many different acquisitions and divestitures.

Unfortunately, their indifference toward UL CA has allowed the problem to remain while Verizon and T-Mobile are often pushing 50-100 Mbps upload over sub-6 with only LTE spectrum. In my area, AT&T maxes out at about 25 Mbps with 15x15 of PCS, while I've been able to achieve 90 Mbps up on Verizon with 20x20 B66 + 10x10 B13 + 5x5 B2.

I really wish that AT&T could work with T-Mobile and Verizon to do more spectrum swaps, even if it were just like-for-like within the PCS and AWS bands. There are a lot of areas where AT&T could have 20x20 MHz or more of contiguous spectrum, if not for a 5x5 or 10x10 slice of spectrum held by T-Mobile or Verizon in between. In many BTAs, simply exchanging specific blocks would give AT&T that continuity across the PCS and AWS bands while preserving or even improving the spectrum holding arrangements for T-Mobile and Verizon, too.

Not only would this help performance, but it would improve mobile battery life, as using fewer aggregated carriers is beneficial from a power consumption perspective.
 
The only thing stopping them from "doing it right" is they do not have enough dedicated 5G spectrum to make it worthwhile in non C-Band areas. Verizon just screwed themselves not being more aggressive with spectrum purchases and being quieter than they should have been about n41 with the T-Mobile/Sprint merger - Verizon totally got "screwed" the most out of that transaction as they *could* have demanded T-Mobile sell off 50% of their PCS or AWS assets in overlap markets and nabbed them up, but chose to stay mostly silent.

T-Mobile "did it right" from the start and just used 600MHz for 5G and aggressively pursued every 600MHz license they could get, and then leased the rest so they have a nationwide decent chunk of low-band 5G spectrum.

In my neighborhood, AT&T has garbage LTE performance due to super fragmented holdings but have a strong C-Band presence (somehow) - so they'd greatly benefit from SA on C-Band around here, and i'd make the carrier's Home 5G options work much, much better here, and not be stuck with sub 5Mbps upload while getting 300-500Mbps download (it's really limited by the ACK packets that can be returned at the moment)

Yeah never understood why they were never really proactive in the spectrum auctions. Is it a matter of saving money? I’m assuming they feel they can make do with their current assets?
 
Verizon really screwed themselves financially and simply did not have the spare cash to spend with the debt load they had after buying out Vodafone's bit.

Then, were forced to act in the C-Band auction and waste 50+bil on that due to T-Mobile's n41 holdings. It's just a string of bad and arrogant management who sat on their laurels too long and by saving 5c upfront to appease shareholders, they ended up spending 10x as much in the end.

And AT&T, this is in my living room attached to C-Band/DoD spectrum, and the speed is so low because the LTE SNR is so dirty on my block upload barely works.

Screenshot-2024-04-19-15-51-19-00-79d0ff8f9752a3fc32487e6d62330a6a.jpg


And they try to sell people on my block 5G Home like it's a suitable replacement with performance like this.
 
Last edited:
Yeah never understood why they were never really proactive in the spectrum auctions. Is it a matter of saving money? I’m assuming they feel they can make do with their current assets?
This was addressed in the mid-2010s: Verizon chose to redirect funding from spectrum acquisition to network densification, which I think was the correct call because density affects network quality much more than raw spectrum bandwidth does.

Now we're seeing AT&T and T-Mobile struggle with uplink performance and downlink consistency because they haven't densified as much as Verizon has — AT&T and T-Mobile's networks have the "speed island" phenomenon much more commonly than Verizon's network does, where their upper midband coverage cells don't overlap, causing speeds to drop precipitously between zones.
 
This was addressed in the mid-2010s: Verizon chose to redirect funding from spectrum acquisition to network densification, which I think was the correct call because density affects network quality much more than raw spectrum bandwidth does.

Now we're seeing AT&T and T-Mobile struggle with uplink performance and downlink consistency because they haven't densified as much as Verizon has — AT&T and T-Mobile's networks have the "speed island" phenomenon much more commonly than Verizon's network does, where their upper midband coverage cells don't overlap, causing speeds to drop precipitously between zones.

Being worried about densification first is also the reason why Verizon is the least consistent performance wise compared to the other two.

It's far more frustrating when using my Verizon phone that it's congested all to hell on one block, and having to go a block over "away from the crowd" to get away from said congestion. I'd rather have the steady speed, even if it's a bit on a slower side like I'd see on AT&T in a congested area... as at least my data will still load.

It was certainly a choice they made, but it's not done any favors for their network quality/perception, as i'm regularly still offering to let my Verizon friends use my hotspot in an area that's severely congested... and even my Visible+ SIM (nowhere near 50GB of use) is regularly encountering data congestion when traveling around for work and doing the LTE/5G/5GUW dance when in hotel rooms leaving me pining for WiFi or using the "other" SIM with me for data that's consistent and usable.

Now, I will say C-Band has helped them a lot - but with how terribly the segregate things like Video Quality, QoS Data, etc.. between "UW" and "Regular" 5G.. it's just not a good look when the phone is consistently swapping between LTE/5G/5GUW
 
Yes VZW is grossly oversold in many markets, to the point that (in the places I go) Visible Depri makes the service literally unusable for many hours, many times per week.

TMO is by far the fastest, hardly ever see the effects of congestion in the places I go

ATT is in between, best geographic coverage in the places I go, not so fast but very reliable.
 
Yes VZW is grossly oversold in many markets, to the point that (in the places I go) Visible Depri makes the service literally unusable for many hours, many times per week.

TMO is by far the fastest, hardly ever see the effects of congestion in the places I go

ATT is in between, best geographic coverage in the places I go, not so fast but very reliable.

Agreed! Although I kind of wish, AT&T would improve their uploads as it’s probably the slowest of the three. Still very usable though.
 
Unfortunate Verizon still provides the best coverage here in the mountains and deserts of Southern California or I would switch in a heartbeat. Also AT&T cannot match my Verizon plan 55+ because they only offer it in Florida. Hmmm, does that mean the everyone over age 55+ lives in Florida? Apparently AT&T thinks so.
 
Back
Top