Cdma

[ads]

Neotel intends to use CDMA-2000, a popular 3G standard in the US. (But not really anywhere else)

Vodacom is starting to sound a bit threatened by Neotel. Interesting.

Here are the facts about CDMA2000, as used by 400 million users in the world, many more than are using Vodacom's flavour of 3G today: http://www.cdg.org/

 

[vodacom3g]

Vodacom is starting to sound a bit threatened by Neotel. Interesting.

Here are the facts about CDMA2000, as used by 400 million users in the world, many more than are using Vodacom's flavour of 3G today: http://www.cdg.org/

Not sure you're comparing apples with apples.

If you want to include 1x in your numbers (which you did and is the vast majority of CDMA2000), you also need to add all the GPRS/EDGE networks in the 3GPP stable.

Maybe rpm can do an article on the relative deployment of the CDMA1/2000 vs. GSM 2G/3G.

We all want Neotel to get going as soon as possible. They've been focusing on their network from the core out, quite correctly so. Thus their first priority is to get the backbones in place and use this to provide additional capacity to corporate users, including networks such as Vodacom and MTN.

Then they'll focus on the last mile and it's here where they're considering EV/DO (the more direct equivalent of UMTS). However, they'll need to build the network and (if feasible) this will take a long time to roll out.

Not knocking Neotel at all (as already stated, we all want them to succeed) but they have a massive disadvantage in developing and rolling out a cell-based radio network. The incumbents have a 10+ year lead on them. Site acquisition alone will set them back many years.

LLU is the answer.....

 

[ic]

As has previously been mentioned, by myself and others, NeeTel is supposed to be the Second National [wired fixed line] Operator as a direct wired fixed line competitor to Telkodemonopolies.

Instead of forcing NeeTel to directly compete with Telkodemonopolies in the consumer wired fixed line market, !CASA decided - after seeing NeeTel's CDMA2000 business plan, to issue NeeTel with a licence and spectrum [800MHz] to operate & provide wireless services to consumers, that means that hell will probably freeze over before Telkodemonopolies is forced to improve its crappy ADSL.

CDMA2000 leaves me stone cold unimpressed, and NeeTel's so called public trial of CDMA2000, seems to have only involved a couple hundred NeeTel employees - it's not about proving that the technology works as it does elsewhere in the world but rather figuring out whether NeeTel is able to supply what consumers want and need in SA, and we do urgently need a wired fixed line competitor to Telkodemonopolies and its crappy ADSL.

I cannot fault NeeTel for wanting to go the wireless route for last mile service delivery to consumers - wireless will still take less time to rollout than a whole wired fixed line infrastructure - but as v3g pointed out, NeeTel still has to rollout all the base-stations and ensure that is has sufficient coverage to cope with demand.

We do urgently need Local Loop Unbundling, sadly guavamentalists like Poison Ivy and Stalin-Mafole continue to protect Telkodemonopolies' wired fixed line monopoly.

 

[Roman4604]

LLU is the answer.....

How so V3G?

By the time LLU is avaliable for use, the market will demand double-digit (Mbps) speeds from fixed-line services. This will require FTTC deployment on a mass scale if one is to cover all major metros & medium towns.

Not a trivial undertaking from a cost & effort point of view. Roll out of a high-speed wireless network to provide the same georaphical coverage if far less daunting in my mind?

 

[ToxicBunny]

Roman : We've been demanding double-digit speeds for nearly 3 years already... do you think its made any difference to Telscum?.. What we demand, and what we get are never the same thing in this country.

LLU is the answer because it will give Neotel access to a completely wired last mile (even though large portions of it need to be upgraded since its so damned ancient)

 

[Roman4604]

Roman : We've been demanding double-digit speeds for nearly 3 years already...

But do you understand why you can't currently have it?

LLU is the answer because it will give Neotel access to a completely wired last mile (even though large portions of it need to be upgraded since its so damned ancient)

Even with access their problem is the "last mile" in this country is actually closer to the last 2.5 miles, and that is where the problem lies.

 

[ToxicBunny]

Yes I understand why we can't have it....Telkom are a bunch of stingy b@!$@#ds who refused to invest in their network when they should have...

With access they'll be able to get people onto a wired infrastructure, and then upgrade it as opposed to building a completely new wired infrastructure including new exchange buildings etc etc etc....

 

[Roman4604]

With access they'll be able to get people onto a wired infrastructure, and then upgrade it as opposed to building a completely new wired infrastructure including new exchange buildings etc etc etc....

I think you do misunderstand, its not the quality of the copper thats the problem, its the length.

Only way around this is to shorten it, which means the exchanges have to move to the street corner, so you will have to build new ones. Not bulidings of course, but mini-DSLAM enclosures, which however require the same level of sophistication (read cost) as a big exchange inc. redundant fibre backhaul, power, ups, enviromental management, vandalism protection etc.

The daunting part is that now instead of having 400 pre-built structures at your disposal, you may need to build many thousands of these new mini-exchanges to get similar coverage as is avaliable now. Even the mighty Telkom with all its money & resources say we may see this around 2011.

 

[Ekhaatvensters]

The network is one and the same ip network. Everything including WiMax goes over the same systems. Voice is essentially VoIP.

They wouldn't be on the same wireless network though, i.e the same towers?

I was thinking that the CDMA will be used by far more people, simply for telephones or internet too, so they couldn't have uncapped connections running off those towers because they will become oversaturated. Or is this not really a problem? I suppose it depends on the extent of their rollout.

I realise that everything will be on the same network once it gets to their internal network, but wireless signal and throughput aren't a problem there.

 

[ToxicBunny]

Roman : Yes I know this is the case.. but having SOME wired network, is better than having none and having access to the Exchange buildings helps reduce the cost of implementing a better last-mile access system. The cost of building the
mini-exchanges" is going to be less than building them AND the main exchanges if LLU doesn't happen.

 

[ic]

IMO Roman4604's point is valid to an extent, but also invalid to an extent, and really Roman4604's point only applies to xDSL and not ordinary analogue telephone lines that NeeTel and others could be taking advantage of if the Local Loop were to be Ubundled anytime soon.

It is very true that the existing local copper loop is on average way too long [copper distance to exchange] and needs to be drastically shortened before ADSL2+ is feasible. The only way of doing this, is to run fibre optic from the exchanges to roadside distribution boxes and put in mini-DSLAMs in the roadside distribution boxes. Telkodemonopolies has been doing exactly that, which is why, in addition to the normal blue & green roadside distribution boxes, you might occasionally see a chunkier white roadside distribution box, these newer white roadside distribution boxes contain mini-DSLAMs, and instead of copper from the roadside distribution box to the exchange, there is fibre optic, copper still runs from the white roadside distribution boxes to each ADSL customer's premises.

As far as ADSL goes, back at the exchange buildings, all the fibre optic cabling from the new white roadside distribution boxes, is aggregated to connect into the core backbone network, so physical exchange buildings are still required to house equipment associated with that aggregation.

IMO Poison Ivy's !CASA effectively stalled LLU by allowing NeeTel to concentrate on [wireless] CDMA2000 as its primary consumer service [which we still don't have], what this means, is that there is no wired fixed line competitor to Telkodemonopolies, to actually drive & accelerate the whole LLU process: if NeeTel was forced to concentrate on a wired fixed line service instead of CDMA2000, then LLU would have been well under way by now, and NeeTel would also be putting in roadside distribution boxes to house mini-DSLAMs with fibre optic back to the exchange buildings that it would share with Telkodemonopolies, and possibly other competitors as well.

 

[ShaunSA]

IMO we're screwed

See i fixed it for you

 

[Roman4604]

so physical exchange buildings are still required to house equipment associated with that aggregation.

All depends on whether this represents the most efficient/cost-effective way to do it for a particular provider.

For example I would imagine if Vodacom were to partake in LLU, they would aggregate their mini-DSLAMs to their closest cellular base station, not Telkom's exchange buildings.

 

[ic]

All depends on whether this represents the most efficient/cost-effective way to do it for a particular provider.

For example I would imagine if Vodacom were to partake in LLU, they would aggregate their mini-DSLAMs to their closest cellular base station, not Telkom's exchange buildings.

AFAIK, the backhaul links that Vodacom and MTN and CellC rent from Telkodemonopolies, for their base-stations, are all aggregated at Telkodemonopolies exchange buildings, if that is the case, then it probably would not make sense for Vodacom or MTN or even CellC to locate their mini-DSLAMs at the closest base-station unless the bandwidth available at such a base-station was upgraded to cope with the increased bandwidth demand imposed on the base-station's backhaul link by the mini-DSLAMs.

 

[Roman4604]

AFAIK, the backhaul links that Vodacom and MTN and CellC rent from Telkodemonopolies

I dont think any provider would contemplate investing in FTTC until they already have FTTN (Node aka local agg. point or exchange). Thats putting the cart before the horse ... networks are built from the inside out.

 

[ic]

I dont think any provider would contemplate investing in FTTC until they already have FTTN (Node aka local agg. point or exchange). Thats putting the cart before the horse ... networks are built from the inside out.

I don't think you are understanding me.

My point is that there is a potential LLU advantage for Vodacom and MTN and CellC, if exchanges do get unbundled, and Telkodemonopolies' competitors are allowed to share the physical exchange buildings with Telkodemonopolies, then Vodacom and MTN and CellC could negotiate with Telkodemonopolies to buy at least some of the the backhaul links they currently rent from Telkodemonopolies - although it would only make sense to buy existing fibre optic backhaul links from Telkodemonopolies - the copper backhaul links [and I've heard there are a lot of copper backhaul links], would have to be cancelled once fibre optic is in place. The logistics of the whole process, would probably be much easier if exchange buildings were unbundled...

 

[Roman4604]

competitors are allowed to share the physical exchange buildings

Why rent when you already own ... shed/container at base stations.

negotiate with Telkodemonopolies to buy at least some of the the backhaul links they currently rent from Telkodemonopolies

Now why would Telkom sell off their (competative) assets ... if you don't like their backhaul rentals, they'll tell you to go put in your own (as a ECNS license holder).

The tricky bit, and that which will determine if there's benefit to using Telkom's exchange buildings, is whether the fibre Telkom lays for its own use (their mini-DSLAMs) between the curb and their exchanges falls within the scope of LLU regulation? I suspect only copper is included?

If it isn't then there's little advantage to using their exchange buildings if you have to lay your own fibre to the curb and you already own alternate aggregation facilites (which may be closer to the curb anyway). And if you already have self-provisioned backhaul there (as Vodacom & MTN are in the process of doing) it makes even less sense to bother with Telkom's exchanges.

 

[ic]

Why rent when you already own ... shed/container at base stations.

Maybe I'm not understanding what you meant by the highlighted bit below, please clarify: did you mean aggregate the fibre optic links from mini-DSLAMs to base-stations, or did you mean actually housing mini-DSLAMs at the base-station sites?

For example I would imagine if Vodacom were to partake in LLU, they would aggregate their mini-DSLAMs to their closest cellular base station, not Telkom's exchange buildings.

 

[Roman4604]

Maybe I'm not understanding what you meant by the highlighted bit below, please clarify: did you mean aggregate the fibre optic links from mini-DSLAMs to base-stations, or did you mean actually housing mini-DSLAMs at the base-station sites?

No I meant just aggregation i.e. locate MetroEthernet switch terminating fibres from mini-DSLAMs at base station. Mini-DSLAMs themselves will have to be at the curbs as that is where the copper from the customers routes to.

 

[ic]

Mini-DSLAMs themselves will have to be at the curbs as that is where the copper from the customers routes to.

I agree - the only way to sufficiently shorten the copper loop is to have mini-DSLAMs housed in roadside [curb] distribution boxes - base-stations would be too far away - just like exchange buildings are [currently on average] too far away, however rolling out HSPA does require more base-stations than 2G requires, so it might be possible in some areas to have mini-DSLAMs housed at base-stations, but is there DSLAM rack space at base-station sites?