Just so I have my information correct

Not quite true.

It cannot be done as "normal practise" but if you are prepared to pay ALL costs involved which might require extra poles and ducting then yes it can, provided there is capacity in the cabinet and on the cable from the exchange.

VM provide links and leased lines etc from BT exchanges. No need for Sky to move.

You'd have to ask for a "network rearrangement", and you might have to pay for the survey for them to even work out what the cost would be.

Don't expect that the cost would be restricted to the 150m between cabinets either. You'd almost certainly need a whole new D-side cable all the way to the house, as I'm sure BT would want to keep the old & new topology separate.

I'm still convinced the marking aren't for FTTC. The cab would be set back in the greenery, and wouldn't need markings on path or road for the cab itself.

What was the cab's status within BDUK/Wales?

Oh, and did BT give a reason for rejecting your friend's offer of £30k?

VM provide leased lines? Is this a business product?

I'd always though VM was FTTC with a coax 'last mile'

In reply to a post by lewisskinner:

VM provide leased lines? Is this a business product?

They've done for a long time - its the business division formerly known as 'NTL business' now known as 'Virgin Media Business'. They can even sell business leased lines in towns and areas where there is no VM cable in the ground.

They even sell leased line connectivity to cellular telcos such as EE. You couldn't sell a business product over a congested shared consumer service.

Edited by jchamier (Sat 25-May-13 22:12:16)

So they're Openreach's Leased lines?

Since Virgin Media have a telecommunications licence, they are allowed to dig the street and run their own in - so no, their VM lines. Cable & Wireless also do the same thing (they're now part of Vodafone group).

In reply to a post by pcoventry76:

They would ruin that greenery to put cement down? I thought the pavement would have been easier?

They tend not to, to make as many people happy as possible, because the pavement is for pedestrians. Obviously there are PCPs on/in pavements everywhere (your one was sort of in a recess in front of some railings, wasn't it?); but I suppose that's either because the pavement was built around it, and/or because back a few decades, there wasn't so much street furniture and so the laws were much more relaxed. A phoneline was obviously considered a necessity, i.e. for anyone to dial 999 but also for people such as doctors who were on call, etc. So that's why the PCPs are/were "allowed" to be there.

With all of the "pavement PCPs" in my town, which have an FTTC twin, they're all across/down the road in a grass verge except one; where there is no level ground (apart from the pavement) that isn't in somebody's garden within 100m. Obviously with cabinets on road junctions, visibility for drivers from every direction takes priority.

Edited by deleted (Sun 26-May-13 01:44:37)

In reply to a post by pcoventry76:

They would ruin that greenery to put cement down? I thought the pavement would have been easier?

Here's a couple where I used to live. Recent housing development (10-15 years old), so are the most recent PCPs added to the exchange anyway, and they had oddles of choice about where to put the FTTC cabinet.

Cab 43, FTTC cabinet across the road, in greenery

Unnumbered cab, FTTC cabinet across road (which need digging), in greenery

In reply to a post by ryant704:

So they're Openreach's Leased lines?

Both. One community FTTH project has VM backhaul over Openreach services. They did a pilot in Cornwall or similar with Vtesse too.

What did the Welsh BDUK project say?

In reply to a post by pcoventry76:

Most of the current EE cells are linked to 1gbps VM lines. Some even have 10gbps when needed.

what's an EE cell ? no way a mobile base station needs anywhere near that.

I think in some cases the cost would be minimal, eg. I have a cab across the road that shares the same ducting as the D side feeding my pole. the problem is BT are very strict, probably for the reason if they do it for one person then word spreads and suddenly everone wants the same treatment and then the cost may add up when rerouting many lines to nearer cabinets, sadly the network design of BTs local loop is far from optimal. A cabinets coverage probably isnt typically a perfect circle round it. I expect its more like a stretched D shape with the coverage on the side furthest away fromt he exchange. Whilst the exchange side of the cabinet probably covered by the previous cabinet.

In my experience of mapping cabinets coverage, that is generally how it works, although the area covered can be any shape, but the cabinet is nearly always towards the exchange side of the area.

in my area its the opposite.

First it needs to be seen to be believed for the line routing from the exchange its unreal, its more than just indirect.

So the routing from my exchange comes from the left side of my front door. My cabinet is to the left at end of road. The nearer cabinet is across the road and the properites behind it are further away from the exchange and they the ones served by that cabinet.

I suppose we agree on the D shape but just in your case the D is facing the other way.

In reply to a post by yarwell:

In reply to a post by pcoventry76:

Most of the current EE cells are linked to 1gbps VM lines. Some even have 10gbps when needed.

what's an EE cell ? no way a mobile base station needs anywhere near that.

Virgin Media are providing a 1 Gbps synchronous Ethernet mobile backhaul service for EE and 3 although probably not from each base station but the local/regional aggregation nodes.
2011 press release

In reply to a post by yarwell:

what's an EE cell ? no way a mobile base station needs anywhere near that.

You sure? I'd have said 1gbps is required, and 10gbps would be in the planning for future stages.

Assume a single cell site; DC-HSPA around 20mbps per user; or LTE at 30mbps per user, and 250 concurrent users = 250 x 30 = 7500 mbps backhaul. (and I gather 250 users on a mast is quite low).

Its people like O2 and Vodafone who are still using 100meg to the mast, which is why its unusual to see over 1.5mbps on a Vodafone 3G / HSPA connection.

In reply to a post by jchamier:

You sure? I'd have said 1gbps is required, and 10gbps would be in the planning for future stages.

Assume a single cell site; DC-HSPA around 20mbps per user; or LTE at 30mbps per user, and 250 concurrent users = 250 x 30 = 7500 mbps backhaul. (and I gather 250 users on a mast is quite low).

HSPA and LTE aren't dedicated bandwidth. They're a group of channels split between everyone on the cell site and contended. There's nowhere on DC-HSPA in the UK that will deliver 100Mbps on a single cell, let alone 20Mb to 250 users on a cell. 3 I believe max out at 42Mbps per cell.

Even the LTE networks don't have enough capacity to allow a single user to hit 100Mb even if they're the only person on the cell. A single STM-1, 155Mb, is ample for a mobile cell. The bottleneck is going to be on the RF between phones and cell, not the backhaul.

If a single cell site could deliver 30Mb to 250 users simultaneously there would be no need for fixed line broadband.

Edited by deleted (Mon 27-May-13 11:53:17)

Openreach are busy providing 1Gbit ethernet circuits for VF and O2. 1Gbit links have been pretty standard for MBNL for some time.

In reply to a post by Ignitionnet:

If a single cell site could deliver 30Mb to 250 users simultaneously there would be no need for fixed line broadband.

So why are the mobile telco's busy signing up backhaul companies to install 1gig+ links?

20 Mhz frequency allocation * 4 bits/Hz = 80 Mbits/s total available bandwidth. LTE data.

If 1 Gig circuits cost about the same to provide as 100M then their would be no point in picking the small one, but your estimates are way OTT even for LTE.

Plenty of mobile base stations have modest microwave backhaul out of town.

So why are the mobile telco's busy signing up backhaul companies to install 1gig+ links?

I guess "are they" is the first question.

The referenced EE / VM thing talks of a gigabit sync-E network (backbone ?) but says nothing of base station capacity.

As the man said, the MHz available to each operator is small and the multiplier to turn that into bandwidth depends on the technology but isn't a large number. "Realistic sector throughput" is the statistic of interest.

BT's mobile backhaul has some standards / interfacing issues that sync-E should sidestep "the service bandwidths which BT Wholesale delivers with MEAS currently are
65Mbit/s to sites connected with 100Mbit/s Openreach Ethernet access circuits, and
300Mbit/s to sites connected with 1Gbit/s Openreach Ethernet access circuits." - OFCOM.

In reply to a post by yarwell:

20 Mhz frequency allocation * 4 bits/Hz = 80 Mbits/s total available bandwidth. LTE data.

If 1 Gig circuits cost about the same to provide as 100M then their would be no point in picking the small one, but your estimates are way OTT even for LTE.

Okay, I was going by expectation rather than actual knowledge. The people I know in the US on their LTE systems can have 5 or 6 people in a meeting room all using LTE dongles in laptops and getting 40megabit+ each - so they must have quite a different amount of MHz. :-/

Plenty of mobile base stations have modest microwave backhaul out of town.

That might explain why in north east hants, the Vodafone and O2 networks generally manage about 1.0 megabit and often less, and the two EE brands and Three can manage 8megabit in the same locations. Perhaps microwave capacity isn't enough for this decade?

There are some interesting calculations on this page on LTE capacity.

For 20MHz of spectrum, it estimates 298 Mbps in 4x4 MIMO. It also includes graphs constructed from tables in the 3GPP specs: 75Mbps in SISO, 149Mbps in 2x2 MIMO.

But those are peaks. The average for the cell, when it includes trickier users out near the borders of the cell, is going to be a lot lower. I think the going expectation (for 2x2 MIMO) is 1.8bits/Hz, making downstream about 36Mbps per 20MHz cell (with 18Mbps on the upstream 20MHz).

Microwave can happily do 1Gb and upwards however if an operator is paying for an E3 that's what they'll get.

Not surprised. Cost difference between 1Gb and 100Mb tail is tiny, no point in taking 100Mb really now. LTE + 3G will potentially saturate a 100Mb circuit even though it would be unusual so better safe.

Saying that though with the network sharing operators are doing when you aggregate 2 physical networks together on the same cell that'd have a fair chance of spending some time over 100Mb.

I can comment on the EE side of things, every base station is being upgraded to 1Gbit capable backhaul when provisioned with 4G services, (some cells are being upgraged eariler where the 3G demand warrants it)

In reply to a post by bookey:

I can comment on the EE side of things, every base station is being upgraded to 1Gbit capable backhaul when provisioned with 4G services, (some cells are being upgraged eariler where the 3G demand warrants it)

The key word there being 'capable' I presume? The cost difference between 100Mb and 1Gb bearers is tiny. The base stations will almost certainly be rate limited to a lower level to avoid wasting cash.

In reply to a post by jchamier:

Okay, I was going by expectation rather than actual knowledge. The people I know in the US on their LTE systems can have 5 or 6 people in a meeting room all using LTE dongles in laptops and getting 40megabit+ each - so they must have quite a different amount of MHz.

Unlikely, the specs don't cover that many options for higher MHz.

Is this 40 megabit the willy waving number reported by the dongle software, or a sustained throughput download - and if the latter, how many of them can get it at once ?

An LTE vendor published a white paper that says 40 Mbits/s is a reasonable maximum sector throughput including the distant clients as well as the locals and a with realistic MIMO configuration for a handset and 2 x 20 MHz - our 4G auction sold off 5 and 10 MHz paired channels in the "go further" 800 MHz band.

Microwave backhaul is scalable and can outstrip a single operator's base station backhaul requirements. It has range and line of sight issues etc but there are many dishes evident on masts and rooftop arrays.

In reply to a post by yarwell:

In reply to a post by jchamier:

Okay, I was going by expectation rather than actual knowledge. The people I know in the US on their LTE systems can have 5 or 6 people in a meeting room all using LTE dongles in laptops and getting 40megabit+ each - so they must have quite a different amount of MHz.

Unlikely, the specs don't cover that many options for higher MHz.

Yes, surely in a meeting room with such required specs, the dongles might well be LTE-capable, but surely they'd be using the company's local WiFi in a meeting? I certainly wouldn't want to be in charge of a meeting worth a lot of time and money, which relied on everyone having adequate LTE connectivity at the same time.

If they all simply show as "connected at 40+Mb/sec", and it is LTE, then that's not really all of them having 40+Mb/sec at once, i.e. one bloke will be doing nothing, one browsing, the next one reading something, and maybe one person streaming a YouTube video in their coffee break. That equals about 0.5Mb/sec of data transfer between the lot of them.

Edited by deleted (Tue 28-May-13 15:56:48)

In reply to a post by yarwell:

Unlikely, the specs don't cover that many options for higher MHz.

Is this 40 megabit the willy waving number reported by the dongle software, or a sustained throughput download - and if the latter, how many of them can get it at once ?

Good questions - my contact was (IT server) technical so I'm pretty sure it was peak throughtput on downloads. At our company we use the same (strange) VPN software that can show the actual throughput being achieved.

An LTE vendor published a white paper that says 40 Mbits/s is a reasonable maximum sector throughput including the distant clients as well as the locals and a with realistic MIMO configuration for a handset and 2 x 20 MHz - our 4G auction sold off 5 and 10 MHz paired channels in the "go further" 800 MHz band.

Okay, so the idea that LTE will give people real-speeds that are faster than the 10megabit peak we get today on 3G/HSPA is a dream once we have more than 2 people using it.

Microwave backhaul is scalable and can outstrip a single operator's base station backhaul requirements. It has range and line of sight issues etc but there are many dishes evident on masts and rooftop arrays.

Makes sense - I guess the masts are grouped, one central has the fibre backhaul and the rest are microwaved to the central ?

Edited by jchamier (Tue 28-May-13 19:10:08)

Some useful real-world test data from the USA.

The RootMetrics app is available on Android and iPhone here in the UK too, but I don't see any reports yet.
http://gigaom.com/2012/04/14/solving-the-lte-puzzle-...

and some info on the spectrum auctions and who bought what here in the UK:
http://www.trefor.net/2013/05/28/analysis-of-who-bou...

Edited by jchamier (Tue 28-May-13 19:29:20)