ajm8127

T-Mobile B4 goes farther than that limit stated for Sprint. I can get Tmo B4 LTE with a -115dBM RSRP at a local hilltop from a site in Sturgis, 22 air miles away. I can even pull 5-6Mbps from it. B12 from the same site is -110dBm. At another location, I can get B12 from Wanblee, SD in Wall, SD...a distance of 36 miles. But that's a -130 RSRP and will drop if you try to use it.

 

Using Tapatalk on Note 8.0

Might be different if you hand into it. It has nothing to do with the band and everything to do with the speed of light/electromagnetic waves. There has to be sufficient "dead" time built into the protocol for random access. If the UE is already synchronized, things are different. I think LTE is the same for all carriers per the 3GPP standards.

That said in rural areas with wide cell spacing, the preamble format can be changed from 0 to something that allows random access from a distance greater than 14 km.

In Sprint's network this would almost certainly be something the market engineer would have to optimize. Most LTE parameters are set in a "golden" template.

The phone is pretty much juts a pawn in this scenario, relaying on direction from the site. However, the phone does have to tell the site what bands it supports, but this is not an easy thing for the user to manipulate. Personally I would think you would want users to attach to the bands in the following order based on congestion and RF conditions.

1. B41
2. B25
3. B26

This has to do with capacity and coverage. B41 coverage is smallest, but has the highest capacity. B26 is used in cases where the other two bands are not favorable, usually for RF reasons, but it could also be because B26 is less utilized (usually only in extreme congestion cases on B41 and B25).

I would suspect your site is just not optimized. I also think if the B25 and B41 PCIs are the same its a relatively safe assumption those signals are coming from the same sector on your nearby site.

There are two modes for handover. Idle and active. Idle mode occurs when you do not have an active RRC connection. Phones will lose their active connections due to inactivity to free up resources and save battery life. They are still "connected" to the serving LTE carrier (what you see in signal check/debug screens), just not active. Active mode occurs when you are transferring data, and for a short time afterwards, which is configured at the site. You can see your connection state at the top of the signal check screen and in the debug screen.

The Spark icon seems to be an indicator of connection state as well. It is greyed out when the connection is idle, white while the connection is active, and spinning while transferring data. This may be specific to my LG G3.

In idle mode the serving LTE carrier will tell the phone which bands to prefer (cell re-selection priority) and how to measure for them. The phone will then act on this information accordingly.

In active mode, the serving LTE carrier will tell the phone what to measure, and it will also have a list of neighbors (cell relations). When certain signal level criteria are met (events A3, A4 or A5 generally which could be based on RSRP or RSRQ) the phone will generate measurements reports and then the serving LTE cell will decide what to do about mobility.

Basically mobility is based on conditions dictated by the site, and then measured for by the phone. It can be based on signal level (more common from my experience) or signal quality.

It might be worth it to call Sprint about this. They are setting up a lot of B41 sites and TDD is new to many market engineers. I would mention that B41 is difficult to connect to despite very slows B25 speeds when outside. Also mention that toggling airplane mode in the main beam of the sector does not get you on B41 (as long as you are sure you are in the main beam, B25 SNR can be a clue). This could be congestion or RF condition based, but probably is just an optimization problem.

actually I edited this... I'd rather keep it away from online. Sorry about that.

 

I will say south of lebanon

That's accurate enough. Thanks.

Just came home for the holidays. Noticed a significant improvement in coverage. My house has an airave because we used to have 1 bar of 3g... But now I am full bars LTE at home on b25.

 

Guess we don't need the airave anymore

 

Sent from my XT1575 using Tapatalk

Do you mind sharing your approximate location such as a nearby intersection? I'm glad to hear progress is being made.

Problem I see you having is the tower is 10.6 miles away. That is about 17 km. LTE uses a cyclic prefix and guard time to allow UEs close the the antenna and far away to synchronize to the site during random access (connecting for the first time) and not interfere with each other.

http://lteuniversity.com/get_trained/expert_opinion1/b/hongyanlei/archive/2010/12/21/cell-size-configuration-in-random-access-procedure-i-preamble-format.aspx

If the preamble format is set to 0, this means the maximum cell radius will be about 14 km. I have only ever seen format 0 used on Sprint sites. If you only wanted to do voice, you would probably be fine, but LTE would likely not work.

Sprint antennas are cross polarized +/- 45° from horizontal from my experience. For the donor antenna, use a yagi/LPDA which supports your bands and has a narrow beam width (I'd say 60° or less). I have mostly seen them mounted vertically.

Place the donor antenna on the roof for isolation from the serving antenna. To further ensure isolation, place the donor antenna on the side of the structure closest to the site. Insufficient isolation leads to the serving antenna signal feeding back into the donor antenna. Think of a microphone in front of a PA speaker. I would put the serving antenna on the ground floor in the center of the structure.

Best bet on choosing a device is calling the company and asking if the specific frequency bands you need are supported.

[EDIT]

Just looked at the weBoost 3G and unless there is a second PCS LTE carrier, you will be out of luck. The PCS range is 1850 - 1990 MHz. G block down link is 1990 to 1995 MHz. The 800 range is 824 - 894 MHz so the down link frequency is covered (866.3 MHz for 8763) but the up link on 800 is lower than that (821.3 MHz for 26763). I am assuming your market uses the same eARFCN for 800 LTE as mine does.

http://niviuk.free.fr/lte_band.php

The weBoost 4G covers 1850 - 1995 MHz, but the 800 range is the same so it would work only for G block.

I know this does not solve the cell radius problem, but is relevant information nonetheless.

Oh, I forgot to mention that for Samsung markets (Cleveland's anyway), the last two digits in the GCI denote the sector: 00, 01, 02 normally or 03, 04, 05 if connected to a 2nd carrier, if available.  Most of the HEX BIDs are 3 digits with the last digit denoting the sector: 1, 2, 3.

This is true of GCIs reported by SignalCheck in general. The last two digits of the 8 digit hex number are the relative cell identity. The first six digits are the eNB ID.

For instance 0x0FF50C3A is eND 1045772 and cell 58. 0x0FE03C19 is eNB 1040444 and cell 25.

The GCI that signal check reports can be combined with the PLMN ID to create the eCGI which uniquely identifies an LTE cell in the network.

Software limits the hardware to two.

Definitely a more concise way to put it.

Android gets confused at times due to downlink and uplink being on the same frequency for TDD with the difference being time allocated for DL or UL. So when it's doing one or the other it tends to put a null value since it doesn't know what to report for EARFCN since both are the same and one is inactive. 

 

Also 41176 is typically the 2nd B41 carrier EARFCN. 40978 would be the first so off the bat it looks like 2 Band 41 carriers live if the shentel site follows the standard Sprint pattern. 

I can confirm those are the two carriers on the nearby site.

I am pretty sure the licenses exist to have five carriers in the Harrisburg/York area. I believe the hardware is limited to two right now. Regardless the bandwidth is going from 30 to 70 overnight with two TDD carriers. They will soon be aggregated so things are going to be interesting when RootMetrics comes back around 1H 2016.

Fortunately Shentel is expanding it's fiber network into PA, so it will be a lot more cost effective for them to deliver the big FTTT backhaul to support the throughput on the air link.

Found this today at 39.97188, -76.78632. I was able to pull 35 Mbps consistently from it. This is the first B41 in Shentel's network as far as I know. I am sure more will follow very soon!

What is up with that UL Channel? That must be a bug. The phone is a Samsung S5.

AJM,

 

Do you know if Shentel will have to sacrifice some CDMA to accommodate the larger 10x10 carrier, or do they have enough spectrum to just add to it without sacrificing CDMA carriers?

 

Sent from my Note 4

I think the PCS A and B bands are 15 MHz wide each. If 10 MHz is devoted to LTE, the remaining bandwidth can support exactly four CDMA carriers. However, I believe there are guard bands so realistically this is only three CDMA carriers. If I had to guess I would say they are using one PCS voice carrier and two EVDO carriers plus the 800 MHz voice carrier.

Theoretically iirc it's 36.7 dl for a 5 mhz FDD and 72.2 dl for 10 mhz FDD using 2x2 at UE.

 

Sent from my Nexus 5X

Thanks for the clarification.

5x5 carrier: 7.44 down, 5.75 up

10x10 carrier: 18.39 down, 18.85 up.

Nice. A 10 MHz channel should theoretically peak at about 60 and a 5 MHz one at around 30. The highest I've seen on a 5 MHz channel is 27ish so I would expect peak in-the-field throughputs of around 54 for the 10 MHz channel. You'd have to in an extremely good SINR area to see speeds like that though and of course loading matters too. 

In Carlisle PA - not familiar with your guys area but showing EARFCN of 8334. Didn't see that reported on here.

That is a 5 MHz LTE channel in the PCS B band. Carlisle is in the Philadelphia MTA, so the channels are different than Woodstock which is PCS A in the BAWA MTA. 8158 is the normal PCA A band 5 MHz channel and 8133 is the 10 MHz channel.  I haven't heard of any PCS B band 10 MHz channel yet for Shentel.

Winchester has been brutally bad lately. I had to switch to 3G at the Apple Blossom Mall to get usable data.

 

Edit: I have an appt in Winchester tomorrow so I can check a couple more towers up there.

 

Sent from my Note 4

I'm surprised about the Mall. Shentel has a small cell in there near their (Sprint branded) store.

No more 10 MHz FDD except Woodstock for now. That is brand new this week. First ALU equipment to do it I think.

That's pretty cool. Is this the first known tower to have 10x10 in the Shentel metropolitan area?

 

Sent from my Note 4

Between this one and the one behind Lowes, yes.

Expect more, especially in Winchester.

My engineering screen says "BW:10MHz"

 

Does that mean 10 total or 10x10?

 

Sent from my Note 4

It's still FDD, so that's 10x10.

The normal second PCS dlEARFCN for Shentel in the BAWA MTA is 8158. This tower is broadcasting 8133. Also based on the speeds on that carrier, it cannot be 5x5 anymore.

2500 MHz 2 carrier aggregation in Harrisburg, York, and Hanover is about to become reality. You guys have noted the 2.5 tower top equipment, but Shentel is placing equipment on the ground now as well. Should only be a couple days away at this point. Probably next week.

GalaxyGuy, you need to do some speed tests on the second PCS carrier in Woodstock near the Spring House. There is something interesting going on at dlEARFCN 8133.

Already being done in at least one market with G block being canablized for a 10 MHz carrier and setting up an adjacent 5 mhz carrier for 15 mhz (5+10) contigious C+G. 

I forgot about 15 being an option. We will have to see what happens.

I could have swore that in the Shentel Conference Call yesterday they said they are now free and clear to proceed with B41 in their entire coverage area.  And that they intend to proceed initially with 125 sites, with many being online before the end of 2015.

The acquisition of nTelos means Sprint  now has the 2.5 spectrum from Winchester to Harrisonburg which nTelos formerly owned. Roanoke is in there too it seems

From here: http://reboot.fcc.gov/reform/systems/spectrum-dashboard

Market - Frequency - Call Sign

• Roanoke, VA - 2496-2690 MHz - B376
• Charlottesville, VA - 2496-2690 MHz -WMH388
• Harrisonburg, VA - 2496-2690 MHz - WMI916
• Harrisonburg, VA - 2496-2690 MHz - WMX327
• Harrisonburg, VA - 2496-2690 MHz - WMX366
• Charlottesville, VA - 2496-2690 MHz - B075
• Staunton-Waynesboro, VA - 2496-2690 MHz - B430
• Winchester, VA - 2496-2690 MHz - B479
• Charlottesville, VA - 2496-2690 MHz - WLW840
• Charlottesville, VA - 2496-2690 MHz - WNTH948
• Harrisonburg, VA  -2496-2690 MHz - WMX331

   

So now that is out of the way, B41 can be deployed in the legacy Shentel network at large as soon as Clearwire is done transmitting (November I believe).

As for Morgantown, Clearwire seems to have the BRS spectrum in Preston and Monongalia counties. So it looks good to go there for the future.

.

Does anyone know if Shentel will be able to add more B25 LTE capacity with the spectrum gained from the purchase of ntelos? I know we already have two 5x5 carriers in B25, but will there now be room for a third 5x5, or possibly a 10x10? If so, that would almost push off the need for B41 in many areas.

 

Sent from my Note 4.

A wider channel like a 10x10 in the PCS A and B Block has come up. I have heard nTelos was using the PCS B Block in the BAWA MTA and Shentel/Sprint was using the A band. If Shentel could re-purpose B Block in that area, they can gain a 10x10 channel immediately, however, much more important is the 2.5 spectrum Sprint picked up in this deal. nTelos owned 2.5 BRS spectrum from Winchester south to Harrisonburg. Shentel will be able to use that as well.

As I type this is see nTelos has 10x10 already adjacent the the G Block. Ultimately there are limitations, one is the devices, but another is the base station. You can only have 3 different FDD LTE channels on a single ALU eNodeB (eCCM2 with three bCEM cards). Right now I think Shentel has a 5x5 800, a 5x5 A or B Block (depending on MTA) and a 5x5 G Block carrier. Adding a 10x10 carrier means removing a 5x5.

I am sure it is still to early for even people within Shentel to know what is going on. The spectrum exchange has to be approved before the deal can close and Shentel can use the acquired spectrum. I expect 2.5 to be deployed on at least a few sites before that happens.

There is lots to think about for the Shentel engineers.

So, are you saying that Shentel's fiber business will likely supply the backhaul?

 

Sent from my HTC One M9 on Tapatalk

Absolutely!

Fortunately microwave can support up to 10gbps, more than enough for B41. But the site it's chained to will require some beastly backhaul.

 

Sent from my XT1030 using Tapatalk

B41 is coming. From some of the conversations I have had I gather Shetel has been given the OK to start putting 2.5 equpiment on their towers. I still do not think they can legally transmit yet. I think it has to do with Clearwire still being on-air until November.

Maybe they are trying more than one type to see what works best.  This location appeared to be a fairly recent install.

What you found was a DAS fed by a nearby macro site.

Any info you're free to share regarding what's happening where will be wildly appreciated! I'm thinking of timelines for specific areas - C'ville (my personal fav), etc.

 

Maybe on one of the sponsor threads?

 

Sent from Tapatalk on either my HTC One M9 or Tab4 10.1, it kinda depends

I can tell you this annoucement was made to the company and the public at about the same time. There is no timeline yet though there will be signifigant discussion and planning before the deal offically closes so the deployment goes as quickly and smoothly as possible.