Network Vision/LTE - Chicago Market

[mhammett]

I have seen times when limited backhaul can cause the SNR to fall with all the retransmits that take place. That was on Evdo. You are far more experienced with the rf side than I am so what would cause snr to fall in the PCS band during the day only?

 

Backhaul and SNR are not related. SNR means Signal to Noise Ratio. It is the number of dB stronger the signal is than the noise. Nothing to do with backhaul. You have the cause and effect wrong. Radio retransmits would happen because of a low SNR. Low SNR is not caused by retransmits. TCP or user-based retransmits, though could be caused by limited backhaul or core capacity or...

 

Not until all sites are live 4G with fiber back haul, we won't know if it is in fact a capacity issue with the back haul or the limited spectrum.

 

From my personal experience in San Diego, when a site does go live with 4G and a neighboring site is not live with 4G it does get congested. It will provide 4G coverage for a larger area until neighboring sites does go live.

 

Microwave can serve just as good instead of fiber.

 

Can a site be LTE active and not have an upgraded fiber backhaul? Maybe that would explain some terrible speeds. Or, maybe sprint isn't delivering as high of a speed line to the sites that they're leading us to believe.

 

It can be active without upgraded fiber backhaul if it is NV microwave. If it is legacy, then no.  The only remote possibility I could see is having LTE on with legacy backhaul is for limited testing.

 

I don't think so, I think Sprint must have the new backhaul in place in order to active the LTE site.  I think this is the reason those folks in Cyrstal Lake were talking about seeing the NV hardware in place for a year on their towers while they waited out the backhaul delay in Microwave towers before it finally went live.

 

and DeKalb ;-)  In DeKalb anyway, all gear was in place (800 panels, 800 MHz radios, 1900 MHz radios, microwave dishes, etc.) Just not active.

 

It's not just during the day and whatever this systematic issue we have been speaking about is not just a Loop or city thing.  I went to the south suburbs on Sunday, drove from the Loop down i-57, and tested speed and LTE signal/SNR values the whole way.  Nothing good to report.  SNR was better, not in the 0-2 range, more in the 6-18 range, but speeds were awful.  I tested while on the road and while I was stationary in the South Burbs.  Was connected to strong LTE signals the whole time and never saw above 500K on either LTE or eHRPD.  Most of the LTE download speeds were between 200-400K consistently....on a Sunday evening in the burbs.  Yikes. 

 

Still can't figure it out, something just isn't right here.  It was just a couple months ago everywhere I went my LTE connection was pulling down 2, 5, 7, sometimes even 12MB/s download speeds, day or night, crowded city spots or not.  Its like it just gone.  I think the SNR values are certainly pointing us to the culprit, whatever that may be, but there is something amiss here at the macro level of Sprint's LTE network across the Chicago area.  I mean for Christ sake is 1:15am Monday morning, I'm sitting in my place in Lincoln Park, and thrilled to see a whopping 370K download speed (and 5.6 SNR) with LTE.  That ain't right!

 

Low SNR is on the radio interface only, nothing to do with backhauls, cores, etc. Usage (or likely low modulation rates for live users) can be responsible for a low of slowness. Low SNR can be caused by faulty transmitters, faulty receivers, faulty sync or outside interference.

 

Here is who has what blocks across the country:  https://wireless2.fcc.gov/UlsEntry/attachments/attachmentViewRD.jsp?applType=search&fileKey=1382746707&attachmentKey=18830422&attachmentInd=applAttach

 

To restate that instead of alphabetical order, but what blocks are adjacent in terms of frequency:

 

PCS A (1850–1865MHz and 1930–1945MHz)- AT&T

PCS D (1865–1870MHz and 1945–1950MHz)- Sprint

PCS B (1870–1885MHz and 1950–1965MHz) Verizon\US-Cellular (Now Verizon\Sprint) - I believe it was originally PrimeCo

PCS E (1885–1890MHz and 1965–1970MHz)- Sprint

PCS F (1890–1895MHz and 1970–1975MHz)- AT&T

PCS C (1895–1910MHz and 1975–1990MHz)- T-Mobile

PCS G (1910–1915MHz and 1990–1995MHz)- Sprint

 

LTE is on G. G is near T-Mobile. Maybe T-Mobile is transmitting out of band more than they should? There is 15 MHz between the top of G in one direction and the bottom of A in the other direction (using FDD). I don't know what occupies that space. Also, in my industry, a strong, nearly colocated transmitter could be even further away than 15 MHz and still cause interference. I assume the cellular guys have better out of band\channel rejection than our gear, but I don't know that for a fact. If so, AT&T another possible source of interference.

[mhammett]

Sorry if this is a dumb question...  but what has changed on the retiring USCC network that it now has problems interfering with Sprint?  Did Sprint push a PRL update to USCC phones to share networks all of a sudden?

 

USCC is very unlikely to be causing LTE issues. They may be causing issues on 1x or EVDO, though. I don't know what Sprint is using for what where.

[Cubbiefan82]

What is sad, is I wouldn't put it pass att to cause problems for other carriers sadly. They will do anything to make others look bad.

[Vince]

Well, T-Mobile is on rapid Chicago deployment.

[cdk]

The mystery device apparently causing the USCC issues is apparently a batch of Galaxy S3s....

 

http://teamuscellular.com/Forum/blog/2/entry-434-certain-device-blamed-for-network-issues/?st=40

http://teamuscellular.com/Forum/topic/5621-galaxy-s3-pulled-off-shelves-today/

 

Now is that problem bleeding over to the Sprint Network (as apparently some of the same frequencies are being broadcasted in the area), or are some of the S3s on the Sprint network creating the problems described here. Or is it could still be one of the other reasons cited earlier....

 

[tommym65]

The mystery device apparently causing the USCC issues is apparently a batch of Galaxy S3s....

 

http://teamuscellular.com/Forum/blog/2/entry-434-certain-device-blamed-for-network-issues/?st=40

http://teamuscellular.com/Forum/topic/5621-galaxy-s3-pulled-off-shelves-today/

 

Now is that problem bleeding over to the Sprint Network (as apparently some of the same frequencies are being broadcasted in the area), or are some of the S3s on the Sprint network creating the problems described here. Or is it could still be one of the other reasons cited earlier....

 

 

Or is it some combination of factors including (but not necessarily limited to) all of the above.

 

A Premier thread suggested that Verizon is also having GS3 problems, but there has been no further discussion or verification of that since Saturday.

 

EDITED:  And it wouldn't necessarily have to be the same frequencies: A defective radio could bleed noise into adjacent frequencies, and as mhammett pointed out, USCC, Verizon, and Sprint are adjacent in the PCS 1900 spectrum in Chicagoland, at least until USCC shuts down here early next year.

 

Your emoticon sums it up perfectly!

[cdk]

Or is it some combination of factors including (but not necessarily limited to) all of the above.

 

A Premier thread suggested that Verizon is also having GS3 problems, but there has been no further discussion or verification of that since Saturday.

 

Your emoticon sums it up perfectly!

 

Possibly!!!

 

The fact that Verizon is apparently having problems with the GS3 suggest that the device could be a problem for any carrier with it (or at least any CDMA carriers). It will be interesting to see this story develop.

[Joeynach]

Possibly!!!

 

The fact that Verizon is apparently having problems with the GS3 suggest that the device could be a problem for any carrier with it (or at least any CDMA carriers). It will be interesting to see this story develop.

This is what I was going to ask.  Why just US Cellular GS3's, why not Sprint and Verizon then.  But also why just recently, the GS3's have been around for 18 months and 6 months ago were superseded by the GS4 (meaning who still buys the older crappier model?).  I didn't start noticing severe network degradation & low SNR until late Sept.  Why now, what changed?

 

Also, why is this happening in just Chicago.  I also don't get that.  In the past few months I have been to Wash DC, NYC, Philly, Atlanta and every time when using LTE got actual LTE speeds.

[Joeynach]

The mystery device apparently causing the USCC issues is apparently a batch of Galaxy S3s....

 

http://teamuscellular.com/Forum/blog/2/entry-434-certain-device-blamed-for-network-issues/?st=40

http://teamuscellular.com/Forum/topic/5621-galaxy-s3-pulled-off-shelves-today/

 

Now is that problem bleeding over to the Sprint Network (as apparently some of the same frequencies are being broadcasted in the area), or are some of the S3s on the Sprint network creating the problems described here. Or is it could still be one of the other reasons cited earlier....

 

And wow this is an amazing discovery, a possible magic bullet variety.  One poster on the forum said, "As of the most current company wide communication. US Cellular will proactively be contacting S3 owners and swapping out for an S4 at no cost to the customer."

[cdk]

This is what I was going to ask. Why just US Cellular GS3's, why not Sprint and Verizon then. But also why just recently, the GS3's have been around for 18 months and 6 months ago were superseded by the GS4 (meaning who still buys the older crappier model?). I didn't start noticing severe network degradation & low SNR until late Sept. Why now, what changed?

There could certainly be issues with Sprint and Verizon models. Why now, best guess a minor revision in a recent production run has caused the problem. Who is still buying them, people who want free phones (or warranty replacements).

[Cubbiefan82]

As crazy as this is, Even though I love my s3. I would be more then happy to trade it in for s4 or even better a new tri-band phone .

[Vince]

This is what I was going to ask. Why just US Cellular GS3's, why not Sprint and Verizon then. But also why just recently, the GS3's have been around for 18 months and 6 months ago were superseded by the GS4 (meaning who still buys the older crappier model?). I didn't start noticing severe network degradation & low SNR until late Sept. Why now, what changed?

 

Also, why is this happening in just Chicago. I also don't get that. In the past few months I have been to Wash DC, NYC, Philly, Atlanta and every time when using LTE got actual LTE speeds.

People still buy older phones. My friends mom just picked up a sprint gs3 2 weeks ago. They're $100 cheaper than a 4, and if someone doesn't need the latest and greatest, it's a viable option. So yes, people still buy them.

[mhammett]

I have a Verizon GS3, got it maybe two months ago.

 

Sent from my EVO using Tapatalk

 

 

[tommym65]

I have a Verizon GS3, got it maybe two months ago.Sent from my EVO using Tapatalk

And does it show any signs of the Great LTE Madness?

[dParis]

I there an explanation in layman's terms out there which explains how a certain handset can foul up an area's LTE network? Is it possible that an update was issued in August/September which triggered the problem?

[mhammett]

And does it show any signs of the Great LTE Madness?

Its hard telling because in most areas most of the time, the VZ network just smokes Sprint's. Verizon's is getting worse while Sprint's is mostly getting better. I have noticed problems with my GPS not working well lately. Open air, fresh AGPS and it can't sync shit.

 

VZ's LTE is really far away, though, so either it would only affect 3G or its such a loud and wideband problem that it reaches all the way down there. That would cause s lot of problems for a lot of non-mobile wireless operations, though.

 

I there an explanation in layman's terms out there which explains how a certain handset can foul up an area's LTE network? Is it possible that an update was issued in August/September which triggered the problem?

Bad components or assembly. Bad firmware. I'd assume it just produces a bunch of noise.

 

Sent from my EVO using Tapatalk

 

 

[Joeynach]

Its hard telling because in most areas most of the time, the VZ network just smokes Sprint's. Verizon's is getting worse while Sprint's is mostly getting better. I have noticed problems with my GPS not working well lately. Open air, fresh AGPS and it can't sync shit.

 

VZ's LTE is really far away, though, so either it would only affect 3G or its such a loud and wideband problem that it reaches all the way down there. That would cause s lot of problems for a lot of non-mobile wireless operations, though.

 

Bad components or assembly. Bad firmware. I'd assume it just produces a bunch of noise.

 

Sent from my EVO using Tapatalk

I wonder if we can pinpoint the time when the LTE network started degrading to a time when a new iteration of the GS3 hit store shelves and/or were moved at heavy volume during a promotion.  

[Joeynach]

As crazy as this is, Even though I love my s3. I would be more then happy to trade it in for s4 or even better a new tri-band phone .

For some reason I really don't think Sprint is going to do what USCC did with the whole S3 for S4 swap.  Sprint might take the S3's off the shelves, and maybe offer some incentives or promotions to S4's from S3's customers, but my gut tells me they can't afford to do this free swap thing.  Nor would they probably think they have to, USCC isn't a major carrier, they need to retain all the customer's they can get, I doubt Sprint would follow suit with something the 6th or 7th largest carrier think that they needed to do.

[Joeynach]

There could certainly be issues with Sprint and Verizon models. Why now, best guess a minor revision in a recent production run has caused the problem. Who is still buying them, people who want free phones (or warranty replacements)

Are we are the conclusion on this board that unless Sprint and their customers 100% remove the use of GS3's from their network than the network will continue to show degraded performance and thus these dial up like LTE speeds?  Because arriving at this conclusion is a huge game changer as far as I am concerned.  Brings up all sorts of issues of what will Sprint actually do or acknowledge (if anything), what will customer's do, and should we consider leaving Sprint if this is going to be a continued issue.

[mc_gusto]

An android blog needs to publicize the usc and possible sprint s3 connection. Maybe even a write-up on the wall here. Then sprint might react

 

Sent from my SPH-L710 using Tapatalk

 

 

[Fraydog]

An android blog needs to publicize the usc and possible sprint s3 connection. Maybe even a write-up on the wall here. Then sprint might react

 

Sent from my SPH-L710 using Tapatalk

 

We could do that here at S4GRU.

 

Problem is, I suspect there just isn't enough evidence yet. Someone like AJ could already be on the case, I don't know for sure but he might be one person here who knows enough to see if a connection is valid. 

 

You can hypothesize something but if you can't prove it, what's the point of running with it? 

[mc_gusto]

It's true too early for sprint. But if the usc problem gets more pub maybe sprint will take note anyway. The usc thing is not too early because they pulled them from shelves. It's speculation but does deserve some attention.

 

Sent from my SPH-L710 using Tapatalk

 

 

[Joeynach]

On a non network degradation note, couple Loop towers now showing 4G upgrades accepted.  One on Wells/Adams and another Wacker/Monroe. 

[Joeynach]

We could do that here at S4GRU.

 

Problem is, I suspect there just isn't enough evidence yet. Someone like AJ could already be on the case, I don't know for sure but he might be one person here who knows enough to see if a connection is valid. 

 

You can hypothesize something but if you can't prove it, what's the point of running with it? 

For what it's worth I have shared our observation with @Sprintcare on Twitter and they told me they kicked it up to supervisors.

[Vince]

For what it's worth I have shared our observation with @Sprintcare on Twitter and they told me they kicked it up to supervisors.

Thank you, Joeynach for your efforts. Someone needs to let them know, and they need network engineers on this 24/7 until resolved. I would love to help, but I have no patience for customer care and their canned responses.