Bell LTE- stable or not?

[technik84]

After a few days on LTE now I'm finding it pretty nice overall. I live in a pretty rural area to the north of Guelph and get a fairly consistent LTE signal. Might only be 1 or 2 bars (out of 5) at times but it's stable and appears to regularly go towards 25-30Mbit download and upload. This is above the estimates given on the Bell site.

I travel a fair amount around the GTA and the signal seems fair. Usually my route is Highway 6 between Guelph and Hamilton. Drops back to H+ along a fair amount of the route through the more rural areas like Puslinch but the coverage map says to expect that. Got signal ok in Mississauga. Relatives in Stoney Creek (Eastern Hamilton) seemed to be on the edge. Sometimes it was LTE and sometimes it kicked back to H+. The map as a whole seems to be somewhat "optimistic". If you're near the edge of the advertised coverage area you might find you're actually beyond it by a few kilometres. Give yourself a bit of leeway.

[frankie5string]

All this talk of certain Samsung handsets having issues with LTE on Bell reminds me of the late 90s and early 2000s when Samsung began making widely available CDMA handsets. The GSM handsets had no real issues. On the other hand the Samsung handsets were generally unreliable and unstable.

Even two handsets of the same ilk can have different issues. We just swapped two Rogers (yes, infallible Rogers) Rockethubs at work because one of them absolutely refused to provide internet. It allowed phone calls but just would not allow internet. All we did was swap the uSIM into the new device and it worked. Rogers thinks the controller was faulty. We tried everything, new firmware, new provisioning, new settings on the unit, initialization and nothing could get it to work.

Given the stability I've had with "rocketsticks" on Bell LTE standing beside unstable Samsung handsets, I'm going to chalk this one up to a revisiting of poor quality radio software in the Samsung handsets. Another reason why I'd never buy a Samsung... It's like 2001 all over again!!

Another thing to keep in mind is that LTE will breath as more people log in. Coverage will grow and shrink throughout the day if you're in a marginal location to begin with. HSPA fall back is there to cover the gaps as the network is built out.

[Rincontrx650]

Thanks all for your input, I am not a cel expert but i do expect to get what I am paying for call me old scholl if you wish but that's just me.
It seems that Samsung's dropping LTE service and not switching to H+ is a common experience I am glad for
'Crikswarren post that I am not the only one finding Bell reps response to shut LTE off as their solution strange...
Even worse is to see Bell's new add for the olympic showing a person with a samsung LTE phone enjoying live TV in an elevator proclaiming this to be the benefits of LTE service... I have a window office in downtown Ottawa and if I step 10 feet from the window my phone drops the connection...

Although I have found the perfect solution for on the road I bought a Wilson signal booster and now get cel service at my cottage in the bush in Norther Quebec go figure.

I guess Viviene's explanation holds true buy a phone with a proven track record.

Cheers.

[VivienM]

I guess Viviene's explanation holds true buy a phone with a proven track record.

This is one reason I like iPhones. Apple makes one or two models per year, they run the same software worldwide, and tens of millions of identical units are sold. So a) they'll be better tested, and b) if there are bugs, there's going to be a LOT of people affected.

Whereas with Samsung, they make... many... models of phones each year, and I think just about all of them run a custom software image for each carrier. So IMO there's a greater likelihood of bugs...

[Steve Punter]

This isn't a slight against Bell Mobility and how it has chosen to implement LTE, but rather as an explanation of your problem.

To allow themselves to expand LTE quickly into more markets than Rogers, Bell has chosen to implement LTE only on SOME of its existing sites. This creates a coverage footprint for LTE that doesn't at all mirror the footprint of HSPA. There are countless dead spots created by a lack of LTE site density that results in ALL of their LTE phones loosing LTE and falling back to HSPA as you drive around.

This isn't specifically a problem endemic to Samsung phones (or, as one troll tried to imply, with all non-Apple products), but rather an issue with the manner in which Bell Mobility has chosen to spend their LTE budget.

While Rogers presently has LTE only in the GTA and in Ottawa in the province of Ontario, their implementation within those cities is 100% (meaning that each and every site that carries HSPA now also carries LTE, thus ensuring an LTE footprint that is more or less identical to that of HSPA). Both providers probably started out with a similar budget that allowed for the installation of LTE on pretty much the same number of sites around the province, but Bell simply chose to make theirs "wide and thin". This should clear up as Bell begins the process of filling in the missing LTE sites.

[WorldIRC]

This isn't a slight against Bell Mobility and how it has chosen to implement LTE, but rather as an explanation of your problem.

To allow themselves to expand LTE quickly into more markets than Rogers, Bell has chosen to implement LTE only on SOME of its existing sites. This creates a coverage footprint for LTE that doesn't at all mirror the footprint of HSPA. There are countless dead spots created by a lack of LTE site density that results in ALL of their LTE phones loosing LTE and falling back to HSPA as you drive around.

This isn't specifically a problem endemic to Samsung phones (or, as one troll tried to imply, with all non-Apple products), but rather an issue with the manner in which Bell Mobility has chosen to spend their LTE budget.

While Rogers presently has LTE only in the GTA and in Ottawa in the province of Ontario, their implementation within those cities is 100% (meaning that each and every site that carries HSPA now also carries LTE, thus ensuring an LTE footprint that is more or less identical to that of HSPA). Both providers probably started out with a similar budget that allowed for the installation of LTE on pretty much the same number of sites around the province, but Bell simply chose to make theirs "wide and thin". This should clear up as Bell begins the process of filling in the missing LTE sites.

I think Qualcomm / phone manufacturers would be benefited if they designed their hardware to operate on the basis that "fallback may be necessary". The phones don't seem optimized for this handoff process yet.

[frankie5string]

I think Qualcomm / phone manufacturers would be benefited if they designed their hardware to operate on the basis that "fallback may be necessary". The phones don't seem optimized for this handoff process yet.

It's not Qualcomm's problem. It's the handset maker's problem. If they chose to use the simplest software from Qualcomm to program their baseband, they'll miss out on some fall back features. If they choose to go deeper and take advantage of more of the chipset radio functions, they'll get more out of it. Exactly the same thing that made early Samsung CDMA phones garbage.

[frankie5string]

While Rogers presently has LTE only in the GTA and in Ottawa in the province of Ontario, their implementation within those cities is 100% (meaning that each and every site that carries HSPA now also carries LTE, thus ensuring an LTE footprint that is more or less identical to that of HSPA).

This isn't true at all. Not every Rogers site has LTE. They do have some LTE-only sites running and pointing into residential areas. Likewise, not every GPRS site has HSPA.

[Steve Punter]

This isn't true at all. Not every Rogers site has LTE. They do have some LTE-only sites running and pointing into residential areas. Likewise, not every GPRS site has HSPA.

You know who you're talking to right? I've been tracking and mapping cell sites for 25 years. I know where the LTE sites are and I'm afraid sir it is you who are mistaken.

Sent from my SGH-I727R using Tapatalk 2

[VivienM]

This isn't specifically a problem endemic to Samsung phones (or, as one troll tried to imply, with all non-Apple products), but rather an issue with the manner in which Bell Mobility has chosen to spend their LTE budget.

Thanks a lot for mischaracterizing what I said. My point was that a company that does one or two devices/year is more likely to properly and thoroughly test their code and their hardware than a company that splits their development efforts on 40-80 devices/year, or however many Samsung makes. If that makes me a troll, so be it.

In any event, isn't this Samsung's problem for selling a device that's unable to gracefully transition between HSPA and LTE? LTE is shiny and new, if Samsung designed their device on the assumption that there would be outstanding LTE coverage, then that was clearly a mistake...

[Steve Punter]

I can't speak for the S3, because I don't own one, but my S2 LTE running baseband UCLF6 works without flaw on Rogers LTE in the GTA. It maintains LTE virtually all of the time. If it does kick down to HSPA it quickly returns to LTE UNLESS there is an active data session.

Sent from my SGH-I727R using Tapatalk 2

[frankie5string]

You know who you're talking to right? I've been tracking and mapping cell sites for 25 years. I know where the LTE sites are and I'm afraid sir it is you who are mistaken.

Of course, we've chatted over the years. That doesn't mean you're always right. Just because a tower has HSPA or LTE doesn't mean it's actually broadcasting. I'd be very surprised if every single Rogers tower in North York or downtown Toronto had LTE running from it. In my case, in North York, they aren't.

If I'm on Rogers, and I'm standing under a Rogers tower, and all I get is HSPA+ with a couple of Rogers Rocket sticks, that's a pretty good indication that LTE just isn't running there. You can tell me there's LTE on every tower until you're blue in the face, and ride your bike to every tower and predict it's carrying LTE, but if my LTE cellphone or rocket stick can't connect to it, it's not all that useful.

At any rate, I agree with the "thinness" comment you had about Bell to some degree - but my experience doesn't tell me the entire Bell LTE network is "thin".

I believe the baseband programming for Samsung handsets makes them poor performers - they always have been. So, go ahead, buy a Samsung LTE phone and enjoy lousy LTE, and then enjoy the lousy fall-back. Then go and get another LTE handset, by a different brand, with the same chipset and see that the network is actually fairly stable. It's like drying clothes made with Dacron in a dryer on high and complaining that the dryer is the reason why they shrivelled to pieces.

[Steve Punter]

Okay, let's see if I can explain this so everyone here can understand:

On a Samsung phone you can go to a ServiceMode screen that gives you information about the current signal. When in LTE this screen shows, among other things, three important pieces of information:

RSRP (Reference Signal Receive Power)
RSRQ (Reference Signal Receive Quality)
PCI (Physical Cell Identifier)

Most important of these for identifying the source of a signal is PCI, which is rather like the LTE equivalent of PN Offset in CDMA. There are 504 cell identifiers in LTE and each sector of each site is assigned one that must not match any of its neighbors. Because there are only a finite number of these identifiers, they are re-used throughout the network, but within a given geographic area they MUST all be unique.

Over the last 6 months I have PERSONALLY visited EACH AND EVERY Rogers site in Mississauga and Brampton, and many of the sites in west Etobicoke. When in close proximity to these sites I have confirmed first that the signal is strong enough to have originated from the site (by looking at RSRP) and then I take note of the PCI for each of the 3 sectors on the site.

I have accumulated this PCI information and found that Rogers follows a strict pattern when assigning PCIs to the various sectors of each site. Each sector is 8 higher than the previous one, and the first PCI never repeats the PCIs used by the second or third sectors in any other site (and so on). So for example a site might have PCIs of 2, 10, and 18, while another might have 3, 11, and 19, but you'll never find one with 10, 18, and 26, as this violates the rule.

Now the take-away from this is that I HAVE IDENTIFIED the PCIs of each and every LTE site in Mississauga, Brampton, and much of Etobicoke. There are no missing sites, or "holes" in the footprint. Perhaps you can tell me which site it was that you "stood under" and observed your phone on HSPA. I will gladly drive to that site (presumably in North York) and confirm your statement.

You can also check the Industry Canada online database for the location of all the active LTE sites. I did that initially to find out where the sites were and so far I have not failed to find one if it was listed in the database. While I can't speak directly for North York, I've thus far identified 161 LTE sites and that's a statistically relevant number (out of the approximately 630 listed on Industry Canada in the GTA). I find it highly unlikely that the situation is any different in North York than it is in Mississauga, Brampton, or Etobicoke (or countless other sites all over the city that I have spot-checked during my travels). I would seriously suspect there's something wrong with your phone, because I've yet to find ANYWHERE outdoors in the GTA where my S2 LTE does not get a solid LTE signal on Rogers.

I should also note that this conversation has veered serious off-topic, so if you would like to continue this "discussion" we should either start a new thread (preferably in the Rogers forum) or take this to PM (though that would unfair to anyone who MIGHT find this interesting).

[toolcube]

Steve, I find this interesting so if you guys start another thread, that would be good. There are 2 Rogers HSPA sites in Woodbridge that don't appear to have LTE despite their coverage maps showing they do. One is at HWY 7 & HWY 400, near the Colossus movie theatres, and the other is at Jane & Teston (tower is on the north side of Teston in between HWY 400 and Jane). I want to note that I haven't done any signal confirmation using phone service modes, etc. I merely looked at the equipment mounted on these towers and didn't see the typical Rogers LTE equipment found on other Rogers LTE sites. The tower at Teston & Jane doesn't even appear to have been updated with their newer HSPA equipment that I've observed elsewhere...the antennae look old etc. My fiance's Rogers iPhone 4S doesn't even pickup full signal being only a few hundred feet away from this tower...I also had spotty 3G coverage in this area when I was with Rogers.

[technik84]

Okay, let's see if I can explain this so everyone here can understand:

On a Samsung phone you can go to a ServiceMode screen that gives you information about the current signal. When in LTE this screen shows, among other things, three important pieces of information:

RSRP (Reference Signal Receive Power)
RSRQ (Reference Signal Receive Quality)
PCI (Physical Cell Identifier)

Most important of these for identifying the source of a signal is PCI, which is rather like the LTE equivalent of PN Offset in CDMA. There are 504 cell identifiers in LTE and each sector of each site is assigned one that must not match any of its neighbors. Because there are only a finite number of these identifiers, they are re-used throughout the network, but within a given geographic area they MUST all be unique.

Over the last 6 months I have PERSONALLY visited EACH AND EVERY Rogers site in Mississauga and Brampton, and many of the sites in west Etobicoke. When in close proximity to these sites I have confirmed first that the signal is strong enough to have originated from the site (by looking at RSRP) and then I take note of the PCI for each of the 3 sectors on the site.

I have accumulated this PCI information and found that Rogers follows a strict pattern when assigning PCIs to the various sectors of each site. Each sector is 8 higher than the previous one, and the first PCI never repeats the PCIs used by the second or third sectors in any other site (and so on). So for example a site might have PCIs of 2, 10, and 18, while another might have 3, 11, and 19, but you'll never find one with 10, 18, and 26, as this violates the rule.

Now the take-away from this is that I HAVE IDENTIFIED the PCIs of each and every LTE site in Mississauga, Brampton, and much of Etobicoke. There are no missing sites, or "holes" in the footprint. Perhaps you can tell me which site it was that you "stood under" and observed your phone on HSPA. I will gladly drive to that site (presumably in North York) and confirm your statement.

You can also check the Industry Canada online database for the location of all the active LTE sites. I did that initially to find out where the sites were and so far I have not failed to find one if it was listed in the database. While I can't speak directly for North York, I've thus far identified 161 LTE sites and that's a statistically relevant number (out of the approximately 630 listed on Industry Canada in the GTA). I find it highly unlikely that the situation is any different in North York than it is in Mississauga, Brampton, or Etobicoke (or countless other sites all over the city that I have spot-checked during my travels). I would seriously suspect there's something wrong with your phone, because I've yet to find ANYWHERE outdoors in the GTA where my S2 LTE does not get a solid LTE signal on Rogers.

I should also note that this conversation has veered serious off-topic, so if you would like to continue this "discussion" we should either start a new thread (preferably in the Rogers forum) or take this to PM (though that would unfair to anyone who MIGHT find this interesting).

This is interesting information.

Can you comment on what is "better" in regards to these numbers. For example on my phone I see RSRP (dBm) -95, RSRQ (dB) -3 currently. It registers as 2 out of 5 bars on the general signal indicator.

I believe as the RSRP figures get larger (eg -105) the signal is weaker. For RSRQ i'm not sure if it's better to have say -4 or whether -3 is "better" in this instance.