new backhaul level required?

[DaQue]

Now that most all sites will get the fast 2500 TD-LTE plus 800 and 1900 FD-LTE what's the level of backhaul going to be needed now. I'm just wondering if the backhaul they ran with the old level of NV upgrades will still be enough

 

(no topics were derailed in asking this question)

[iansltx]

Fiber-fed sites will be fine. Sprint may need to up their bandwidth transport commits but that's about it.

 

Wireless backhauled sites on the first or second hop away from fiber-fed sites will be fine. They've got a few hundred Mbps of throughput.

 

Wireless fed sites several links down the chain probably won't see TD-LTE, unless they can get fiber pulled closer to them (less towers sharing a single wireless backhaul).

 

That's my best guess anyway. #1 is pretty much a certainty.

[gnoj]

Sprint is requiring at least 100 Mbps backhaul at each NV site, with contracts that allow for future scalability up to a gigabit. From what I understand the great majority of sites are satisfying this requirement with fiber, with only a few microwave-fed. I'm not aware of any places where there is more than one hop between a microwave site and the core site fed with fiber, so backhaul won't be much of an impediment to TD-LTE deployment once NV 1.0 upgrades are finished in that area. If that is an issue somewhere they will address it, since every site will be getting TDD 2600.

 

The theoretical data draw on the tower if all 3 sectors broadcasting Ev-DO and 5x5 PCS LTE have their airlinks saturated does exceed 100 Mbps (about 123 Mbps if my math is correct), although that is only if everyone connecting has an ideal signal. Adding LTE 800 does lift the theoretical draw to about 235 Mbps but the real-world demand should still fall under 100 Mbps. Once more TDD channels are added and linked together through carrier aggregation, Sprint will I'm sure then have to take advantage of those backhaul contracts and turn up the fiber to 200 or higher, which shouldn't take much effort at all.

[Dkoellerwx]

Sprint is requiring at least 100 Mbps backhaul at each NV site, with contracts that allow for future scalability up to a gigabit. From what I understand the great majority of sites are satisfying this requirement with fiber, with only a few microwave-fed. I'm not aware of any places where there is more than one hop between a microwave site and the core site fed with fiber, so backhaul won't be much of an impediment to TD-LTE deployment once NV 1.0 upgrades are finished in that area. If that is an issue somewhere they will address it, since every site will be getting TDD 2600.

 

The Kansas market has some pretty long microwave chains along the highways. Especially between KC and Wichita, between KC and Springfield, and then north to the Iowa border. 4-5 sites chained at least.

[Mobilesolutions]

sprint does not need more than 100mbps at each site yet, unless that site is the master for wireless backhauled sites around it.

[gnoj]

The Kansas market has some pretty long microwave chains along the highways. Especially between KC and Wichita, between KC and Springfield, and then north to the Iowa border. 4-5 sites chained at least.

 

Interesting. That'll probably become more common as they begin to expand the rural footprint. I wonder what the max throughput to that 5th site would be. If it's much less than 200 Mbps they might indeed have to pull fiber a little closer. They'd probably wait until after 2600 is up to see what the load is like first though.

[khammondnm]

Just curious, if fiber is in place, is scaling to a gig really expensive? I don't understand why they wouldn't just go all out. I know there is some differences in the equipment, but I always thought the main cost was the fiber line itself.

 

Sent from my SPH-L900 using Tapatalk 4 Beta

 

 

[Mobilesolutions]

Just curious, if fiber is in place, is scaling to a gig really expensive? I don't understand why they wouldn't just go all out. I know there is some differences in the equipment, but I always thought the main cost was the fiber line itself.

 

Sent from my SPH-L900 using Tapatalk 4 Beta

 

Depends who is providing the circuit, I know i have seen price sheets where 100mbps is $1100 month and a gigabit is $1900.

[z250kid]

Now that most all sites will get the fast 2500 TD-LTE plus 800 and 1900 FD-LTE what's the level of backhaul going to be needed now. I'm just wondering if the backhaul they ran with the old level of NV upgrades will still be enough

 

(no topics were derailed in asking this question)

they can get what ever speeds they like from fiber.   just have to pay a lot. I'm not sure if there going to need more pairs

[digiblur]

On the docs I've seen and dug up around the net.  The router in the cabinets typically have a fiber port capable of 1 gigabit. I'm sure that's pretty scalable as well.  Since it's normal to run multiple pairs and if the provider could only do 1 gig fiber they could just light up another pair and bond them for 2 gigabit.  Small problem to say the least... 

[d_rail34]

On the docs I've seen and dug up around the net.  The router in the cabinets typically have a fiber port capable of 1 gigabit. I'm sure that's pretty scalable as well.  Since it's normal to run multiple pairs and if the provider could only do 1 gig fiber they could just light up another pair and bond them for 2 gigabit.  Small problem to say the least...

 

If I remember correctly, depending on the length/distance of the cable, it is up to 1gbs...but...I think if the the distance is somewhere around 300ft. or less then it's pretty much infinite so long as you have the hardware to support it. Though I highly doubt the latter would be the case.

[digiblur]

If I remember correctly, depending on the length/distance of the cable, it is up to 1gbs...but...I think if the the distance is somewhere around 300ft. or less then it's pretty much infinite so long as you have the hardware to support it. Though I highly doubt the latter would be the case.

 

You are thinking of Cat5 cabling specs.  The specification calls for a 100 meter distance between two devices.  Fiber is a lot different. The distance varies due to the different hardware, cabling, etc.

[d_rail34]

You are thinking of Cat5 cabling specs. The specification calls for a 100 meter distance between two devices. Fiber is a lot different. The distance varies due to the different hardware, cabling, etc.

It has been awhile since I have messed with it, but I was thinking with multifiber it was 1gbs at around 2kilometers, and anything inside of 100m was pretty much limitless. Damn it digi...now I'm gonna have to google it.

 

Oh...and cat5 could only be ran around 400ft. (I think) max, and the data rate was still less than 50mbs...I think...

 

Ok...as soon as I break for lunch I'm googling. Now you have me SERIOUSLY second guessing my memory. Lol

Edited August 1, 2013 by d_rail34

[d_rail34]

Ok...so cat5 specs have changed quite a bit. Mind you, it's been since the early/mid 90's since I installed it. As for the fiber specs, that was accurate. Unless there's something I'm missing as far as type goes.

[ddimit]

Most fiber run to site will either be CWDM or DWDM Muxed.  Telco will provide a pair or single strand of Single Mode Fiber hand off  to provide  either 1 GB connections or 10GB (still quite pricey) to each Colo at that location.  Each wireless provider then does a VPN tunnel over the link back to their Switch center.

 

Microwave is more common than you think., even in Urban areas. Sometimes it not cost effective for the special construction costs to do fiber last mile. Microwave is a pretty quick cost recovery to back haul to a central site versus paying by mile tor MAE(Metro Area Ethernet) to each location.

[maximus1987/lou99]

Most fiber run to site will either be CWDM or DWDM Muxed. Telco will provide a pair or single strand of Single Mode Fiber hand off to provide either 1 GB connections or 10GB (still quite pricey) to each Colo at that location. Each wireless provider then does a VPN tunnel over the link back to their Switch center.

 

Microwave is more common than you think., even in Urban areas. Sometimes it not cost effective for the special construction costs to do fiber last mile. Microwave is a pretty quick cost recovery to back haul to a central site versus paying by mile tor MAE(Metro Area Ethernet) to each location.

Disappointing. I was led to believe all Sprint's 38k sites were fiber to the tower.

[S4GRU]

Disappointing. I was led to believe all Sprint's 38k sites were fiber to the tower.

By whom??

 

Network Vision sites are approximately 80% Fiber/AAV and 20% Microwave. Of those that are AAV, most are fiber fed and just the final connection is by a 3rd Party vendor. Like a cable company.

 

The 38,000 Sprint sites is the entire network. Many sites, like mountaintop sites, or remote highway sites, it makes more sense to go microwave. Seeing how microwave can support 1-3Gbps and you can set up several links side by side if you need more, then what's the big fat hairy deal about microwave backhaul? Nothing. That's what.

 

Robert via Samsung Galaxy Note 8.0 using Tapatalk

 

 

[maximus1987/lou99]

By whom??

 

Network Vision sites are approximately 80% Fiber/AAV and 20% Microwave. Of those that are AAV, most are fiber fed and just the final connection is by a 3rd Party vendor. Like a cable company.

 

The 38,000 Sprint sites is the entire network. Many sites, like mountaintop sites, or remote highway sites, it makes more sense to go microwave. Seeing how microwave can support 1-3Gbps and you can set up several links side by side if you need more, then what's the big fat hairy deal about microwave backhaul? Nothing. That's what.

 

Robert via Samsung Galaxy Note 8.0 using Tapatalk

Then why not microwave more of the network? Save a bunch $$, right?

[S4GRU]

Then why not microwave more of the network? Save a bunch $$, right?

Microwave is not really cheaper. It depends on how close fiber is to the site. Microwave has more up front equipment costs. You pretty much only want to use microwave for sites where you cannot get fiber or fiber backed AAV reasonably. Sometimes it makes sense.

 

For instance, with Clearwire WiMax, microwave was a good solution. Each site only needed 8-12Mbps speeds. So if you have 8 sites connected to a 100Mbps connection, all is good. However, Sprint Network Vision that will not be a good solution with triband LTE where the carriers will be well burdened. It would be fine for low capacity sites, like the GMO sites. But high capacity sites will need an ample fiber connection.

 

Your doom and gloom perspective is just not warranted. If a site gets set up with triband and is fed by AAV or microwave and it turns out to not be enough...do you think that's the end of the road? Sprint is not the company it was. It is now owned by SoftBank...deep pockets and high expectations. The new network is going to be monitored. If a site starts to underperform, its going to be corrected.

 

Also, do not forget that Sprint has scalable backhaul contracts. If they have a 100Mbps connection at a site from Cox cable, and they now need 200, they just call Cox. Cox comes out within 30 days and adds more. Most of these AAV vendors have fiber backed connections. Adding more capacity at these sites is not difficult in most situations. And at a microwave site, they just need to add backhaul capacity at the single donor site, and all the connected sites get a speed boost.

 

This is all good news here.

 

Robert via Samsung Galaxy Note 8.0 using Tapatalk

 

 

[bigsnake49]

Microwave is not really cheaper. It depends on how close fiber is to the site. Microwave has more up front equipment costs. You pretty much only want to use microwave for sites where you cannot get fiber or fiber backed AAV reasonably. Sometimes it makes sense.

 

For instance, with Clearwire WiMax, microwave was a good solution. Each site only needed 8-12Mbps speeds. So if you have 8 sites connected to a 100Mbps connection, all is good. However, Sprint Network Vision that will not be a good solution with triband LTE where the carriers will be well burdened. It would be fine for low capacity sites, like the GMO sites. But high capacity sites will need an ample fiber connection.

 

Your doom and gloom perspective is just not warranted. If a site gets set up with triband and is fed by AAV or microwave and it turns out to not be enough...do you think that's the end of the road? Sprint is not the company it was. It is now owned by SoftBank...deep pockets and high expectations. The new network is going to be monitored. If a site starts to underperform, its going to be corrected.

 

Also, do not forget that Sprint has scalable backhaul contracts. If they have a 100Mbps connection at a site from Cox cable, and they now need 200, they just call Cox. Cox comes out within 30 days and adds more. Most of these AAV vendors have fiber backed connections. Adding more capacity at these sites is not difficult in most situations. And at a microwave site, they just need to add backhaul capacity at the single donor site, and all the connected sites get a speed boost.

 

This is all good news here.

 

Robert via Samsung Galaxy Note 8.0 using Tapatalk

 

They don't need to come out in most cases unless the terminal equipment needs to be changed out. Most of the time it's done on an internal website somewhere.

[WiWavelength]

To ensure that it has sufficient backhaul, Sprint has brought in a consultant, Sir Megs-a-Lot, whose mantra is "Baby Got Backhaul."

 

 

 

AJ

[jefbal99]

To ensure that it has sufficient backhaul, Sprint has brought in a consultant, Sir Megs-a-Lot, whose mantra is "Baby Got Backhaul."

 

 

 

AJ

That is one of my karaoke jams!!!

[WiWavelength]

That is one of my karaoke jams!!!

 

Oh, good, then our S4GRU CES booth karaoke lineup is coming together.

 

Ben has already signed on for "Damn It Feels Good to Be a Gangsta."

 

http://s4gru.com/index.php?/topic/4346-lte-800-deployment-maps/&do=findComment&comment=179403

 

You have "Baby Got Back."

 

And I am sure that Mike, the SignalCheck guy, from New England can be counted on to do "Brady."

 

 

AJ

[mikejeep]

Oh, good, then our S4GRU CES booth karaoke lineup is coming together.

 

Ben has already signed on for "Damn It Feels Good to Be a Gangsta."

 

http://s4gru.com/index.php?/topic/4346-lte-800-deployment-maps/&do=findComment&comment=179403

 

You have "Baby Got Back."

 

And I am sure that Mike, the SignalCheck guy, from New England can be counted on to do "Brady."

 

 

AJ

 

Oh my god I'm going to have to stop my random searches for "signalcheck" if this is what I going to start finding.  Where do you dig this stuff up from?!  Ahahaha..

 

Brady is the man, but I can't say that I am ready to commit to singing that.  Ever. 

 

-Mike

 

EDIT: I just flipped between page 1 and page 2 several times to see what happened.  Still can't figure it out.  Mega-nerdy fiber discussion on page 1, and suddenly it's MTV out of nowhere on page 2.  I love this site.

[mhammett]

Fiber-fed sites will be fine. Sprint may need to up their bandwidth transport commits but that's about it.

 

Wireless backhauled sites on the first or second hop away from fiber-fed sites will be fine. They've got a few hundred Mbps of throughput.

 

Wireless fed sites several links down the chain probably won't see TD-LTE, unless they can get fiber pulled closer to them (less towers sharing a single wireless backhaul).

 

That's my best guess anyway. #1 is pretty much a certainty.

 

All of the microwave licenses I've looked at have had 100 megabit per site serviced from that link. They didn't build out the links to reliably scale higher than that. They didn't use large enough dishes to have sufficient gain to accommodate throughputs that are multiples higher.

On the docs I've seen and dug up around the net.  The router in the cabinets typically have a fiber port capable of 1 gigabit. I'm sure that's pretty scalable as well.  Since it's normal to run multiple pairs and if the provider could only do 1 gig fiber they could just light up another pair and bond them for 2 gigabit.  Small problem to say the least... 

 

They're probably using MPLS or Carrier Ethernet or... to limit each circuit to something less than the line's capacity of X gig. Most are 1 Gig, but 10Gig isn't impossible.

 

Most fiber run to site will either be CWDM or DWDM Muxed.  Telco will provide a pair or single strand of Single Mode Fiber hand off  to provide  either 1 GB connections or 10GB (still quite pricey) to each Colo at that location.  Each wireless provider then does a VPN tunnel over the link back to their Switch center.

 

Microwave is more common than you think., even in Urban areas. Sometimes it not cost effective for the special construction costs to do fiber last mile. Microwave is a pretty quick cost recovery to back haul to a central site versus paying by mile tor MAE(Metro Area Ethernet) to each location.

 

They may or may not be doing a VPN. I'd suspect all of the tunneling is done on the service provider's side. This would allow them to maintain a 1500 byte MTU. Also, Metro Ethernet is often called MetroE.

 

By whom??

 

Network Vision sites are approximately 80% Fiber/AAV and 20% Microwave. Of those that are AAV, most are fiber fed and just the final connection is by a 3rd Party vendor. Like a cable company.

 

The 38,000 Sprint sites is the entire network. Many sites, like mountaintop sites, or remote highway sites, it makes more sense to go microwave. Seeing how microwave can support 1-3Gbps and you can set up several links side by side if you need more, then what's the big fat hairy deal about microwave backhaul? Nothing. That's what.

 

Robert via Samsung Galaxy Note 8.0 using Tapatalk

 

FYI: Even when a cable company is used, its going to be a fiber hand-off. They have very extensive fiber plants to support their HFC (Hybrid Fiber Coax) plants.

 

Then why not microwave more of the network? Save a bunch $$, right?

 

I would have.

 

Microwave is not really cheaper. It depends on how close fiber is to the site. Microwave has more up front equipment costs. You pretty much only want to use microwave for sites where you cannot get fiber or fiber backed AAV reasonably. Sometimes it makes sense.

 

Robert via Samsung Galaxy Note 8.0 using Tapatalk

 

Capital vs. operating. For example, the Zayo pricing I have puts 1 gig at 4x the price as 100 meg, yet obviously 10x the bandwidth. That's a $10k/month difference when looking at 10x 100 meg sites aggregating to fiber vs. 10x 100 meg sites each with their own fiber. The cost of a complete link capable of 500 megabit true throughput (1 gigabit when using various header compression methods on "typical" Internet traffic) is $10k or less. That's a 10 month ROI on the decision to go microwave vs. fiber. That position would only improve when you start exceeding 100 megabit per site as the microwave gear can already handle it (depending on how many sites are in that spur\loop). No more investment needed to realize the reduction in fiber costs.