reedacus25
-
Posts
138 -
Joined
-
Last visited
Content Type
Profiles
Blogs
Articles
Media Demo
Gallery
Downloads
Events
Forums
Everything posted by reedacus25
-
I'd take a WAG that far reaching Verizon roaming, including in-market, is not inexpensive. Couple that with T-Mobile's extremely surgical and conservative roaming footprint and limits, as well as much more inbound roaming usage compared to Sprint since inbound CDMA roaming isn't nearly as plentiful as it is on the GSM side.
-
While Sprint has an existing MPLS and IP network, that's not being sold in the last mile. They still require last mile at minimum in terms of backhaul to reach the core, if not requiring middle mile transport to get to your Sprint POP. Sprint and T-Mobile's backhaul costs should not be all that different. The big cost is in last mile construction and leasing. And middle mile is an aggregate pipe back to the core. T-Mobile and Sprint are likely leading similar pipe sizes. ~500-600 Mbps for a standard 3 sector site with 20 MHz of LTE and probably 2-3 UMTS carriers plus GSM. Sprint with 2x 5 MHz (B25 and B26) plus 2-3 B41 carriers should consume ~600 Mbps on link budget. This is all conservative math, it's likely closer to 700 I'd bet. But they should be roughly the same, other than the fact that T-Mobile didn't incur the same construction costs in the LTE cycle as they moved from TDM to IP in their UMTS cycle for a good majority of sites, MRC should be roughly similar.
-
I'm seeing 20x20 of AWS-1 (D+E+F) to compliment the 25x25 of PCS (E+F+C) and the L700A block in both Broward and Miami-Dade Counties.
-
Or same year launches, if you're lucky. How long has your site sat there converted but offline?
-
Catering to the masses. They get more mileage out of their capacity by allowing high SNR/RSRQ UE use the network. 64QAM pushes more bits in the same resource blocks. So if you still have excess resource blocks to consume, why not let it feed the masses? If a UE needs voice in a low RSRP environment, they shouldn't have any issues, since their voice traffic will be carried over QCI=1, being low bitrate, high priority. Still get your coverage benefits for voice coverage, and you get 25% more data capacity in the LTE RAN. Would you rather serve 200 customers close to the cell and 5 customers on cell edge adequately, or would you rather serve 100 customers close to the cell and 20 customers further from the cell? My numbers are exaggeratory to facilitate the point I'm attempting to make, which is that catering to the masses is more efficient than appeasing the stray few on the fringe. Yes, these are broad sweeping statements, I'm just making a generalization that they can make better spectral use by satisfying the higher signal quality UE in the cell, thus why you're seeing CA.
-
If there are excess resource blocks available on the L700 side, and there is sufficient demand on the network, does it not make sense to fulfill the demand? There is no Band 2 to aggregate with currently, so that's not on the table in this scenario. And clearly this is in a high RSRP/RSRQ area, so 64QAM frames, meaning more mileage out of the network in terms of overall throughput. There's a reason in this example that L700 is the supplemental component carrier. It's resource block allocation is supplementary to the resource block scheduling of the L2100 RAN which is the primary component carrier. So if the packet scheduler needs to allocate resource blocks to L700 for UE not in range of L2100, i.e. Edge of service and indoors, the first non-free resource blocks dynamically allocated are going to come from CA connected UE.
-
I find it disconcerting that you find the voice quality to be subpar with T-Mobile. They have long since outpaced ATT and Verizon in terms of VQ from just about any perspective. I haven't used Sprint for voice in around 10-12 years, and Sprint's voice quality for me was just a hair above T-Mobile's FR AMR VQ on GSM. After moving to UMTS, after it finally made it to me, the bar was raised even higher, and then again with AMR-WB intra-carrier. I've had several close family members use my phone for one reason or another to speak with another family member on the other end of the line only to comment on the 'clarity' compared to their ATT lines. So I'm curious what brought your subpar quality statement out.
-
I'm very interested to see if T-Mobile ever makes a move on Omaha. Obviously getting a local exchange is a big first step, but surely that is achievable, especially with the 531 area code overlay. Surely there's a full prefix they can claim in 531 at a reasonable price. I don't even think 402 is fully exhausted yet either. I can say that I'm sure it works in a very narrow scope. I've had polar opposite experiences based on location. In 2007 when the CWS was at Rosenblatt, service was almost non-existent. One device on our account at the time was a Treo 650 and couldn't hold a connection for anything. Totally shot. Nokia 6101b and 6131 I had at the time worked barely outdoors in the grandstand. Pressbox was zilch, and in Doorly Zoo there was no chance of signal. However, now at TD Ameritrade, signal in the area was just fine in 2013. Thats because there's a site I believe on the Creighton building at the corner of Mike Fahey and N 17th. Coverage in the downtown area at street level was very hit or miss, and popping inside was a death trap for pink airwaves. However, it is totally worth noting that T-Mobile, ironically enough, was the only provider with decent data at TD Ameritrade in 2013, and this was pre-LTE. ATT and Verizon had COWs beyond the left field wall, and Sprint didn't have much in the way of NV yet in the area, and nothing in the vicinity of TD Ameritrade, so all of the CSpire roamers clogged up the 1x channels on Sprint, and their phones were bricks. ATT LTE was B17 only, and was clogged up, as was Verizon, yet T-Mobile's DC-HSPA was humming along (GPRS, not even EDGE on the GSM side). Later we went out towards Boys Town for dinner one night, and phones were dead. CSpire (by way of Sprint) saved us in that jam. Outside the 680 perimeter is no mans land for T-Mobile, and inside it is hit or miss or miss or miss. Eppley was also really bad for signal as well, considering the signal was coming from a site in Council bluffs 3 miles away I guess it was pretty good all things considered. My guess is with Sprint having a decent network, plus US Cellular adding a 4th player in the market, they see the market as too crowded to excel in. But the existing customers coming to visit Omaha are the ones who really suffer without a better network to fall back on.
-
The Metro train was a good one to ride for T-Mobile, in terms of subscribers, spectrum, and DAS infrastructure and certain macro leases and properties. I couldn't find a final subscriber count at the closing of the merger, but most pinned at ~9M. And in July 2015 ahead of the final CDMA sunset, they claimed to have captured 92% of existing CDMA subs and moved them to GSM. They also quoted 500,000 remaining CDMA subscribers after that 92% capture rate. So somewhere between 6.25M-9M came from the metro base, which accounts for roughly 21-30% of the 29M+ that were added in <3Y time, depending on what the actual number was. So there is still an extraordinary surge in subscriber density in tons of markets across the country, contributing to congested air links. Sent from my iPhone using Tapatalk
-
Let's assume a pretty predictable trajectory for network load as subscribers increase. Explosive subscriber growth has caused this runway to the end of the trajectory to become much closer than it was say 1+ years ago. Every time they add more LTE capacity, be it by L700 deployment or UMTS->LTE refarming for wider LTE channel width, it extends that trajectory out further, extending the runway from that point, but not changing the fact that the massive increase in subscriber growth, takes you way further down the trajectory than anything else. For reference, total subscribers after 13Q1 were 33.968M, and after 15Q4/EOY were 63.282M. That's 29.314M subscribers added to the network in 11 quarters/less than 3 years, 86% more subscribers on a network in that amount of time. [All numbers pulled from T-Mobile earnings reports from their investor relations site] Where we are now is the intersection of being "too successful too quick" and being able to put money back into the network with tangible results. Densification, sector adds/splits, all to mitigate the congestion of the massive subscriber growth. It's just not happening quick enough to make everyone happy. Another thing to remember is that it stands to reason that it is likely that a large majority of adds are in urban areas, which were already susceptible to congestion because that's where the majority of the existing subscriber base came from. All this to say, BingeOn likely extends that runway trajectory, but not further than the net adds drags it right back down. Sent from my iPhone using Tapatalk
-
"Settling" isn't in human nature. If I can go to a restaurant and have Kobe beef for hamburgers, regardless of price, but knowing that I can't taste the difference, might not eat all of it, and more importantly, they have a very limited number of those patties, most people are going to say "hell yeah I want the good stuff." Same for a "nice" bottle of wine, if you can't taste the difference. If I can have a Visa Signature or World Elite MasterCard variant of my credit card, even though I may never use those features, and it's going to cause my merchants to pay a higher transaction rate, I'm going to want the "better" version. Verizon has their marketing slogan currently set to "never settle" to play on that. The same logic applies to the "videophile" user base. If I can have a perfectly color corrected reference display with reference level source quality, why wouldn't you? But with mobile video, and data as a whole, you are sharing the resources with everyone, and there is finite capacity. Sure, you pay for your unlimited or X amount of data that you pay for, you want to use it however you see fit. Great. But some people can see it as driving a Hummer for the hell of it. Nothing wrong with Hummers, but they are poorly efficient, take up too much space in parking lots, etc. It looks nice. So if I don't need the Hummer, and a Camry does just fine, why add those carbon emissions and burn up that gas and take up all that space in the parking lot? If I don't actively need 1440p video on my phone, and I can gather just as much information and experience from lesser quality video, why use those network resources unnecessarily? Why not be a good neighbor? Why not use something that I know fits in the overhead of my link budget and will never need to buffer compared to a higher quality/bandwidth/resolution stream. My whole point is that we as people thrive for excess that we don't need, video quality not withstanding. I'm not saying we should watch Braille, but there is absolutely a point of diminishing returns when it comes to a screen that fits in your pocket. Are my analogies perfect? No. Am I attacking you? Not my intention. But if you can be less resource intensive with virtually no perceivable penalty for most people, it creates a better environment for everyone.
-
Well considering that you are scaling at a perfect 3x from 854x480 to 2560x1440, I'd probably say it looks pretty decent considering a linear upscale. But then again I wouldn't call something on a phone a 'reference display' either so I have a lower expectation. My point is that at 480p, and even less, I can resolve all of the detail that I need, for a fraction of the bandwidth. If you want the best for having the bests sake, go right ahead, but I'm not enjoying my video any less because it's not 1:1 pixel mapping. Sent from my iPhone using Tapatalk
-
Just for reference, I think it's kind of silly to insinuate that watching "480p" on a ~6" display at likely anywhere between 8-16" viewing distance would cause significant eye strain or even pain. We all love to be videophiles, and we all want the best possible anything, but we tend to want excess as well. Now this quote may have easily been tongue in cheek, but for example, playing back a DVD rip of 'The Office' from Plex at less than 1 Mbps is more than sufficient to enjoy it. Video statistics are 624x352@24 MP4 codec in an AVI container, average bitrate of 943k. Audio is MP3 container @48kHz, stereo, VBR averaging to 131k. All together <1024Kbps aka 1 Mbps. For me, that is a wholly acceptable amount of detail on a phone. Even mostly acceptable on my 65" DLP. I'm not saying people can't enjoy better bitrates, better quality, better speeds, etc. but often times its diminishing returns. People just always "want the best" when they won't notice a difference between best and decent on screens this small. Just wanted to say that 1.5 Mbps is totally usable for anything I'm trying to do. This is not me advocating for people not having choice for video quality, but rather giving an example of how usable 1.5 Mbps can be more than acceptable for quality on a small screen. Sent from my iPhone using Tapatalk
-
Sad that lots of those L1900 sites are SISO that you've come across. Around these parts they are certainly the exception, not the rule, thankfully. And it also helps that most of those sites are in very off the beaten path locations, so often unburdened as well. Also love the random 10 MHz pockets in some parts. Makes me jealous to see it go so underutilized. Sent from my iPhone using Tapatalk
-
By having vast amounts of spectrum, and generally more dense networks than Sprint and T-Mobile. For example, its not uncommon for T-Mobile to have 10-20 MHz of deployed spectrum for UMTS, as well as 20-50 MHz deployed spectrum for LTE. ATT will typically have 30-70 MHz of deployed LTE, and likely 20-40 MHz of UMTS deployed, coupled with a denser grid of macros.
-
There's a pretty great chance that some or all of the GSM (/GPRS/EDGE) layers in ATT's network in your area may have been sunset already. They have been thinning GSM pretty steadily as they add more LTE carriers. So in this scenario, either your particular serving cell no longer has GSM, or they may be running GSM in PCS only (for whatever reason) where UMTS may be running on 850. One of the reasons GSM typically out "carriers" UMTS signal is that with GSM it's a multi-frequency network, versus single-frequency, so each channel has a 'dedicated' spectrum allocation, allowing them to run GSM at higher power levels than UMTS which has to manage co-channel interference due to its single-frequency design. Obviously with GSM you have a reuse factor, but in most non-urban environments, the network isn't dense enough to need to significantly dial back the GSM side. Sent from my iPhone using Tapatalk
-
Just to clarify, there is expansion happening, on both the 700 spectrum, as well as non 700 spectrum, mostly in Roam Over Build areas. Mostly dictated by if 700 is available to deploy on. So while T-Mobile is expanding some of their footprint filling in some gaps and moving the boundary of the edge of service with L700, there is a ton of expansion in non-overlay projects as well. Sent from my iPhone using Tapatalk
-
I'm not sure I would call densification plans "copying Sprint." All the carriers are densifying to some extent. T-Mobile has been densifying plenty of markets to date, as well as having plans for many more. Their densification strategy in my market has been mixed and varied with multiple approaches. In some areas they have built brand new sites from scratch, in most areas however they are co-locating. With most of these new sites in this market, they have been L2100/U1900 only sites, with no GSM carriers or now sunset U2100 carrier(s). Hell, they've even been adding sectors to some heavy load areas to increase capacity. In some cases they have deployed small cells, or mini macros if we are being honest. In some cases they are full tri-sector builds alone. Some are single or multi-sector builds strung across multiple nodes. Its been very interesting to watch. Some are located on existing utility poles, some had poles built for the purpose. I do also know that they are overlaying Band 12 in Band 2 GMO land. I don't believe these sites are being reworked for tower top radios on the existing PCS side of the radio chain, but they are getting Band 12 layers either GMO or tower top radio'd. I wouldn't count on T-Mobile, or anyone, sitting on their hands in 2016.
-
Just to be clear, my questioning of the results had no correlation to who "won" other than the methodology of the sampling not being clear, with results being given in Bytes per second, versus bits, and my assumption being that if the measurement is in Bytes, then it must be a total throughput over time, which should directly correlate to total traffic measured for each provider. Sprint having the prevailing speeds in the race wasn't what made me question it, it was 2x3=4 that I was confused about. Sent from my iPhone using Tapatalk
-
I still don't get how they came to their conclusion, not that I am looking to you for answers. The DAS operator showed total throughput figures in xTB, and an average throughput in MB/s. We can either accept an off by factor of 8x error for B vs b, or we can assume that this should equate to total average backhaul throughput being metered for each carrier on the backside of the DAS. If the backhaul being metered is the sampling data used for the average throughput metric, I do not see how those figures, over equal time, equal the same total traffic count. I've also seen other, albeit older, DAS architecture in person, and can't imagine how else they would measure individual carrier throughput in that manner considering the radios behind the node antennas have been dumb in the past. RF from BTS to light, back to RF to clean and base line, then to light again for distro, then back to RF from radio to node. More to your post I quoted, I believe everyone is taking a slight angle at the data to fluff their own stats. But I don't think there is enough data from this report for me to take it at face value is all. Need more info and variables to compute "network potential" aka excess speed metric than what is given in the report. But bottom line it sounded like everyone performed mostly well across the board, so yay progress right? Sent from my iPhone using Tapatalk
-
Not old deactivated SIMs, new un-activated SIMs allow network attach and speedtest traffic. Sent from my iPhone using Tapatalk
-
I'd be lying if I said the thought hadn't crossed my mind before. The issue I'm seeing in these Root reports is that its a mostly nationwide trend, spanning vendors. Traditionally I believe NSN markets were marginally better performers in reliability, but that would be anecdotal at best evidence with no real tangible evidence. But even NSN markets are nose diving with Call Blocking. But most of this started with the change to the Note 4 for testing for T-Mobile. And has continued to be an issue in most cases. I mean pretty much everything in the Southeast pretty much rose in call blocking, the state of TN is a prime example. Look at Memphis, Nashville, and Knoxville, near perfect to really poor. Something is aloof, and it sounds like VoLTE is not enabled? Something isn't right here.
-
Whats interesting in the Grand Rapids example is that the call blocking (and dropping) both got better between 1H and 2H. But then there are tons of locations where the call blocking rate "blooms" so to speak at ridiculously high rate, New Orleans being a good example. T-Mobile in Michigan should all be NSN territory, correct. Would be really nice if someone (Root) could compile the data into a CSV breaking down all the metrics per market, per carrier, per testing period, so that we could see period over period trends and change. Would be really helpful in my opinion.
-
How common is this? Reason for my asking is I've used VoLTE nearly ubiquitously since October of last year. There were definitely some hiccups in the first few months, but that seemed to be worked out by the end of the year, and the only real issue was in high mobility, not low mobility scenarios. The only issues I have now are when my iPhone now on iOS 9 decides to poo itself and it nukes IMS registration over LTE or Wifi and I have to toggle airplane mode to remedy it. Even when that happens it falls back to UMTS for voice without much trouble, maybe a 5-10 second delay for it to pull its head out of its rear. The reason I bring this up is because my market, New Orleans, is an area I have had no issues at all with VoLTE, or voice traffic in general, VoLTE or otherwise. Yet in 1H2015 (early March) T-Mobile's voice statistics were 0.1:0.4 blocked:dropped. This was when T-Mobile was still running at 5 MHz wide for their LTE RAN (2.5/2.8 Mbps down/up). Fast forward to today and the LTE RAN has been widened to 10 MHz in the latest testing date (late August) and the speeds increase expectedly to 7.8/10.0 while calls go to 1.8/0.7. In that same time frame, T-Mobile has added over 10 new sites in the area, as well as modernizing remaining legacy sites that got skipped over. My point being, I have used VoLTE and voice extensively with no issues of blocked calls, and have polled other users in my area, both Android and iOS for their feedback, and they too have not had issues with blocked calls, yet RootMetrics reports an 18x increase in blocked call failures, that I have not found. Even when I am in heavy load sectors where I get sub 1 Mbps down even on 15 MHz wide channels, VoLTE never hiccups. I am not saying that the Root testing method is flawed, but I am saying that the fact that Call Blocking is up significantly nationwide after switching devices is highly suspicious. Could the devices not be functioning properly? Who knows, but it plays into the gloom and doom narrative of the "Post BingeOn Age" when I have seen nothing first hand across multiple markets to link these call blocking rates to anything of substance. Sent from my iPhone using Tapatalk
-
GSM sunset is going to be tricky. In dense urban markets, they have all the more reason to sunset it. They are already thinning it to as low as reasonable so as to maximize UMTS capacity in PCS, which frees AWS up for LTE widening in AWS. The master plan. The issue is that in rural areas they have little to no reason to go UMTS (much to my chagrin). Where you need minimum 2.4-4.2 MHz FDD of GSM depending on your reuse factor, in rural areas you can get away with less than 1 MHz on highway 2-sector sites strung down a road way. In some areas, this would require scaling back the LTE carrier to support a UMTS carrier if they chose to go this approach. If I could have 10 MHz Band 2 + GSM or 5 MHz Band 2 + UMTS I'm probably going to settle for the latter. But I am the latter. By doing the GMO quick and dirty overlay, they have boxed themselves in a PCS corner. The other piece of this is that T-Mobile has already shown that they aren't afraid to have LTE only RAN. Look at all the B12 only sites in Michigan, North/South Dakota, Louisiana. They clearly have enough faith in VoLTE and their LTE RAN to not feel like they need the other layers. It will be interesting to see how this all turns out in the long run. Sent from my iPhone using Tapatalk
