43Mbps HSPA+ milestone for Vodacom

[jannievanzyl]

I won't argue with how you have explained the capacity - 43Mbps, it just didn't cross my mind. However, putting it as "nitpicking" is perhaps over sensitive on your part but I understand given the negative feedback VC has been receiving lately.

Yeah, sorry. I've been trying to engage here and staying factual, hoping to engage in proper debate but it seems people have lost the art (of logical debate) and are only out to attack. To be honest, I find it quite offensive if I state verifiable facts and people just openly say I'm lying. I've been on this forum for many years and I'd like to see one example where I intentionally mis-stated factual information.

I understand they're either Cell C fanboys or just hate Vodacom, but surely logic prevails? Guess not.

So, my apologies.

For the important part, I still need your feedback on why these areas have poor services:

1. Pretoria
2. Soweto
3. Johannesburg - Central
4. Sandton
5. Centurion - Central

True, no celco can claim 100% coverage but I believed these areas should be top on your list? Sandton, in particular is your test centre for LTE, surely that should say something.

Of course they are. The way I always answer these "why don't you have coverage in this or that are?" questions is as follows:

"As most people agree the cell companies are only after your money, why would they NOT provide coverage in highly populated areas so they can make even more money off you? Surely it's not good business?"

So something is stopping the operators from deploying networks in those areas and the number 1 reason, by far, is capacity issues. And, again by far, the biggest capacity issue is new towers to deliver that capacity.

Given enough towers (and adequate backhaul), everyone will have decent coverage and throughput. It really is as simple as that; more and more people want to use the network but it's getting more and more difficult to get permission to erect new towers.

Everyone wants coverage but no-one wants a tower in their neigbourhood. So how do you solve this?

There are smaller, secondary reasons why quality might suffer in the short term and one of those are relevant in Gauteng at the moment. Every now and then you need to refresh your networking kit, its a natural cycle in all things technology.

Vodacom embarked on such a refresh quite a while ago (thus you see a 43 towers popping up everywhere) but it's not a quick process. You have to do it in such a way you don't affect the live network (too much). It's kinda like adjusting your car's tappets with the engine running.

After such a refresh, you have to re-optimise the network and this is happening in a number of the areas you've mention. Once done network quality is back or better than before.

But it still does not solve the high-site issue. As always, if your roof is available and you've sorted your neigbours, give us a call!

 

[jannievanzyl]

The FACT that you change parameters to get different coverage maps(predictions) and the FACT that you are getting poor coverage in places that you would think should have great coverage, and the FACT that you used the word "prediction" means that you cant prove what you are saying and we cant disprove what you saying. Hence you can say whatever you want and call it a FACT. For all we know you changed the coverage maps (predictions) for what the perfect scenario for vc is and used that as your marketing or like you are now you are using theoretical numbers...

That's why the invite to you to come and check for yourself.

When are we doing this?

 

[MyLeftFoot]

Then grab yours and divide R185,00 by 500 and you'll also be surprised by the result (37 cents and not 34 cents). Then multiply that 3 cents difference by 500.... wow you have misquoted your prices by a whopping R15,00!!!
It's not painting it RED?
http://www.vodacom.co.za/portal/sit...cted=4_a_Bro_0_Bro_1221&ciFormat=default&ht=t

I wonder how prevalent this kind of thing is. When i was working for a call centre we found that one of our telephony providers was over billing us by R15000-R40000 per month and the only reason we found out was because i was tasked to write a app to read the CDR's we got from the provider and calculate the cost directly from the CDR's.

 

[MyLeftFoot]

That's why the invite to you to come and check for yourself.

When are we doing this?

Man i am not taking the day off of work for you to show me something you can fake.
Did you not get that from my previous post?

 

[marine1]

Yeah, sorry. I've been trying to engage here and staying factual, hoping to engage in proper debate but it seems people have lost the art (of logical debate) and are only out to attack. To be honest, I find it quite offensive if I state verifiable facts and people just openly say I'm lying. I've been on this forum for many years and I'd like to see one example where I intentionally mis-stated factual information.

Do you forget the Iburst tower debacle? Sory to bring this up but.... Maybe not mis-stated factual information but was unethical at least.

I understand they're either Cell C fanboys or just hate Vodacom, but surely logic prevails? Guess not.

So, my apologies.

I hate every SP, you are all in the same boat but at least Cell C seem to be doing more than you guys have done in a long time. You are ripping off the public and you think we are stupid, I dont care about the technical crap that is spoken about. Fix it and lower the prices, stop wasting billions on sport and other sponsorships

Everyone wants coverage but no-one wants a tower in their neigbourhood. So how do you solve this?

Why is MTN better in terms of coverage? I guess you will say Vodacom has more users. I must give Vodacom a thumbs up compared to MTN with service at the call centre, when you can get through, MTN is a complete fail with this.

 

[jannievanzyl]

Yes, it was intentional, you not following the norm (like all other MyBroadbrand users) to confirm a speedtest.
This is also a reason or contributor to the mis-trust being generated by Vodacom lately.

I'm sorry I don't follow the norm. I create screenshots because I use them for different purposes. And some of them are from an Ookla server that does not have the link capability. Or have you not noticed?

As much as you want to try and brand me a liar, I'll stand by my offer for you to come and view anything we do at Vodacom to see for yourself. It's easy to hind behind a forum nick and throw out accusations.

Pop over and come and show me (and VC) where anything I've posted here about coverage, speeds, etc. are false.

Let me know when you want to pop over. As always, the beers are on me.

 

[jannievanzyl]

Man i am not taking the day off of work for you to show me something you can fake.
Did you not get that from my previous post?

What I got is that you will hide behind an anonymous nick but don't have the guts to prove your accusations. Even when you get an open offer for assistance to do so.

At least, that's what I get from your posts.

Even if you work for Cell C, it's OK. I've got nothing to hide.

 

[MagicDude4Eva]

FACT: RED is limp

This is a test at our offices in Bryanston (few hundred meters from Didata campus).

Enough said: Test of Vodacom (iPad) and MTN (iPhone 4)

Test on Vodacom - http://www.speedtest.net/iphone/108106277.png
Test Date: Apr 18, 2011 11:35
Connection Type: Cellular
Server: Midrand
Download: 1.40 Mbps
Upload: 0.30 Mbps
Ping: 246 ms

External IP: 41.30.147.116
Internal IP: 41.30.147.116
Latitude: -26.0456
Longitude: 28.0267

Test on MTN - http://www.speedtest.net/iphone/108106171.png
Test Date: Apr 18, 2011 11:34
Connection Type: Cellular
Server: Midrand
Download: 3.89 Mbps
Upload: 0.11 Mbps
Ping: 160 ms

External IP: 41.121.155.68
Internal IP: 41.121.155.68
Latitude: -26.0445
Longitude: 28.0241

As the ladies will tell you, it's hardly ever about size, but how long you can last (=signal) and how long you can keep it up (= 3G) before you go limp/soft (= Edge).

We threw Vodacom 3G backup devices out, as not even with 3G booster antennas we could achieve a reliable signal. Connecting a MTN/CellC dongle produces speeds in excess of 7Mbps whereas Vodacom bounced between 3G/Edge delivering sub 1Mbps.

So in the middle of Bryanston's IT-hub, Vodacom's radio-engineers must have guest-i-mated those towers wrong and staff with Vodacom contracts have to walk into the carpark to take a call.

 

[Quest10n]

...

Why you taking what I said out of context? Just go check again to what I responded to.
I hope an appology will follow as I never challenged you to proof anything.

 

[g0nt4du]

That's why the invite to you to come and check for yourself.

When are we doing this?

Isn't a solution to this whole debate as follows. What if there was an independent body that simply tested and presented values of real world experience. So far I have seen Vodacom and CellC own up to the fact that the speeds are theoretical and you would get much less in real. that's why the battle spilled over to speedtest.net results where it is apparent today that Vodacom has higher average speeds.

As for people hating and ranting on the forums, yes it is in your rights. But I believe it does not accomplish anything. We have MNP now in the country. No cell provider has any kind of SLA with you. So if possible, move to another network that suits you. that is of course after raising your issue with the network. Although, its not in the network's best interests they can tell you to get lost if you don't like their service.

You see complaining about coverage today, they not going to put up a tower for you tomorrow. I am certain they have stats/analysts looking over what regions they have low subscribers and what regions are high.

We must also understand that sometimes, stupid regulations and red tape actually prevent these networks from being their best. take the case of sentech, they sitting on spectrum and now they dive in. Instead of being more useful. Had they ceded that, then vodacom/mtn, etc could have had access to it, to improve service/capacity. Might not be totally accurate but you get the idea...

Like @jannievanzyl said, lets engage in some constructive debate. Yup he can prove the test methodology to you. But hey not many of us would even understand the technicalities of what he has mentioned. and to the average layman, all u can say is there are lots of factors involved. hence the need for an independent technically competent body to evaluate claims.

So ya, if you don't like the service, complain. If its too bad to live with, ditch it. The networks will get the message. Any competent network, will ensure that they don't get ditched. That's business really.

 

[jannievanzyl]

Wow! It's been exactly two years since I complained about poor networking signal in my area, I usually get GPRS and if I am lucky I get EDGE, if it's done via software how come I am still in a GPRS area? What kind of software is that, tenderpreneur supplied (Gijima)

Well you have painted everything red but it's still slow, in fact even more slower. You can change the frame but the picture remains the same, same old Vodascum! Putting lipstick on a pig?

Man, would we (and any other operator) love to upgrade GPRS to HSPA+ via software. I see a Nobel price coming!

Why don't you post (or PM) me your location and I can check what the story is.

The 2G (i.e. GPRS/EDGE) coverage is ~99% of population today while the 3G is 74%. So there'll be plenty of areas where you'll get 2G but not 3G.

And it'll vary from network to network as not everyone have the same number of towers in the same areas or even have enough backhaul to the towers already there.

 

[jannievanzyl]

What if there was an independent body that simply tested and presented values of real world experience.

Good point. Such a body should be Idependant, know about Communications, be an Authority and based in South Africa.

We'd be very happy to be part of such an effort.

In the UK, OFCOM performs this function.

 

[marine1]

An independent body in SA? bwahahaha

 

[Joe Average]

I'm sorry I don't follow the norm. I create screenshots because I use them for different purposes. And some of them are from an Ookla server that does not have the link capability. Or have you not noticed?

As much as you want to try and brand me a liar, I'll stand by my offer for you to come and view anything we do at Vodacom to see for yourself. It's easy to hind behind a forum nick and throw out accusations.

Pop over and come and show me (and VC) where anything I've posted here about coverage, speeds, etc. are false.

Let me know when you want to pop over. As always, the beers are on me.

I know for a fact that I did not call or brand you a liar. I said it creates suspicion (...small a red rat...) when the norm is not followed. I have noticed the Ookla server, thank you. I did go to the Vodacom shop at a big mall in my area over the weekend - and they could not show me a modem that could run at 43Mbps - and I don't feel like driving to Vodaworld to test it out. I might take you up on your invite....but it will have to be after Easter weekend.

 

[MyLeftFoot]

Isn't a solution to this whole debate as follows. What if there was an independent body that simply tested and presented values of real world experience. So far I have seen Vodacom and CellC own up to the fact that the speeds are theoretical and you would get much less in real. that's why the battle spilled over to speedtest.net results where it is apparent today that Vodacom has higher average speeds.

As for people hating and ranting on the forums, yes it is in your rights. But I believe it does not accomplish anything. We have MNP now in the country. No cell provider has any kind of SLA with you. So if possible, move to another network that suits you. that is of course after raising your issue with the network. Although, its not in the network's best interests they can tell you to get lost if you don't like their service.

You see complaining about coverage today, they not going to put up a tower for you tomorrow. I am certain they have stats/analysts looking over what regions they have low subscribers and what regions are high.

We must also understand that sometimes, stupid regulations and red tape actually prevent these networks from being their best. take the case of sentech, they sitting on spectrum and now they dive in. Instead of being more useful. Had they ceded that, then vodacom/mtn, etc could have had access to it, to improve service/capacity. Might not be totally accurate but you get the idea...

Like @jannievanzyl said, lets engage in some constructive debate. Yup he can prove the test methodology to you. But hey not many of us would even understand the technicalities of what he has mentioned. and to the average layman, all u can say is there are lots of factors involved. hence the need for an independent technically competent body to evaluate claims.

So ya, if you don't like the service, complain. If its too bad to live with, ditch it. The networks will get the message. Any competent network, will ensure that they don't get ditched. That's business really.

+1
completely independent body! No funding from the SP's.

 

[Antonf]

One thing Vodafone SHOULD get credit for is it's coverage map, if not the coverage itself. You can zoom to Fido's kennel, and it is accurate. At my one home the coverage map shows 3G available at the street, but not at the eastern end of the property. I've walked the length of the property with my notebook and watched the signal strength on MDMA. The map is accurate.

 

[Quest10n]

How do you want me to prove it? I'll be happy to accommodate you.

Coverage claims are based on very complex radio planning software that calculated propagation areas and overlays it over population distribution. All the networks use these.

I'll re-extend my offer for you, or anyone else, to come and sit down so we can show you how it's done. I would also suggest you ask the other networks to do the same for you. Then YOU can compare the numbers and report back.

Let me know when you want to do this. I'll immediately arrange it.

For others who won't be able to join Odracir, let me give a brief overview.

If any radio planning guys are here, please give a better explanation. Once Odracir has seen and done it himself, he can come and explain better.

Basically you have a map of the population distribution of South Africa on a geographical map of South Africa. Geography is obviously very important as it affects RF propagation.

You then overlay your towers on this map and give the software a signal level to work against. Most networks use -100dBm for 3G. The software then calculates each towers radiation pattern to the point where the signal level falls below that level. The result is the typical coverage map you see posted by all the operators. By adding up how many of the population falls within the coverage, you can quote a single number. For VC it's over 99% for 2G and 74% for 3G.

You can see if you pick a lower level as the cutoff you'll get a much bigger map and subsequently can quote a much higher %coverage.

But you have to pick a realistic cut-off number, one that'll still give you a decent connection. And you have to take to the number of actual users into consideration. Otherwise you end up with a number that does not reflect the real world.

The end result of picking the wrong numbers is a map that does not bear resemblance to actual experience and you'll quickly get lots of complaints by subscribers who say they're in coverage on the map but cannot connect at all to 3G or get very poor throughputs.

To avoid this, Vodacom always run the prediction on the conservative side, exactly to avoid these type of complaints. I've seen a flood of these complaints recently (not about VC) on the forum and it's the exact result you'll get if you're overly aggressive with radio planning prediction tool.

For example, if you use -115dBm (as some networks do, actually only one I know about), you can either claim a much bigger % coverage, but what's the quality like. Vodacom's 3G coverage of 74% suddenly becomes 90% but who are you fooling? So we stick with the standards and do not try and spin the numbers.

For those who do not know the dB scale as it relates to power output; Every -3dBm is half of the previous level, so going from -100 to -115 is 32-times weaker. You'll agree that's going to make a huge difference. Not 15 times, but 32 times.

The second important point revolves around the 3G phenomena of cell shrinking. This is part of WCDMA and affects all networks, even those ones running on 900MHz.

As soon as you load the cell, it gets smaller. And very quickly. Put a few users on (less than 10) and the cell shrinks to a fraction of what is was unloaded. Again, users will see this suddenly losing 3G when they used to have it. Or very low speeds. (The only way to fix this is by adding more towers, but that's another topic, although it's been discussed here recently)

To get back to radio planning, the question is then; what loading do you use to run your prediction? If you use no loading and a low signal level, you can calculate you're *best* coverage number. But you probably should not as those complaints of lost-3G and slow speeds will hit you again. And you forumites are quick to point them out!

Again, Vodacom is very conservative around this.

:wtf: Why you taking what I said out of context? Just go check again to what I responded to.
I hope an appology will follow as I never challenged you to proof anything.

 

[jannievanzyl]

I know for a fact that I did not call or brand you a liar. I said it creates suspicion (...small a red rat...) when the norm is not followed. I have noticed the Ookla server, thank you. I did go to the Vodacom shop at a big mall in my area over the weekend - and they could not show me a modem that could run at 43Mbps - and I don't feel like driving to Vodaworld to test it out. I might take you up on your invite....but it will have to be after Easter weekend.

Cool, let's do it Joe. I really would prefer it if you guys have the facts. Does not mean you have to agree with the company or what it does, but facts are just that.

And with the correct factual information, you can make an informed decision, whatever that decision is.

 

[Quest10n]

How do you want me to prove it? I'll be happy to accommodate you.

Coverage claims are based on very complex radio planning software that calculated propagation areas and overlays it over population distribution. All the networks use these.

I'll re-extend my offer for you, or anyone else, to come and sit down so we can show you how it's done. I would also suggest you ask the other networks to do the same for you. Then YOU can compare the numbers and report back.

Let me know when you want to do this. I'll immediately arrange it.

For others who won't be able to join Odracir, let me give a brief overview.

If any radio planning guys are here, please give a better explanation. Once Odracir has seen and done it himself, he can come and explain better.

Basically you have a map of the population distribution of South Africa on a geographical map of South Africa. Geography is obviously very important as it affects RF propagation.

You then overlay your towers on this map and give the software a signal level to work against. Most networks use -100dBm for 3G. The software then calculates each towers radiation pattern to the point where the signal level falls below that level. The result is the typical coverage map you see posted by all the operators. By adding up how many of the population falls within the coverage, you can quote a single number. For VC it's over 99% for 2G and 74% for 3G.

You can see if you pick a lower level as the cutoff you'll get a much bigger map and subsequently can quote a much higher %coverage.

But you have to pick a realistic cut-off number, one that'll still give you a decent connection. And you have to take to the number of actual users into consideration. Otherwise you end up with a number that does not reflect the real world.

The end result of picking the wrong numbers is a map that does not bear resemblance to actual experience and you'll quickly get lots of complaints by subscribers who say they're in coverage on the map but cannot connect at all to 3G or get very poor throughputs.

To avoid this, Vodacom always run the prediction on the conservative side, exactly to avoid these type of complaints. I've seen a flood of these complaints recently (not about VC) on the forum and it's the exact result you'll get if you're overly aggressive with radio planning prediction tool.

For example, if you use -115dBm (as some networks do, actually only one I know about), you can either claim a much bigger % coverage, but what's the quality like. Vodacom's 3G coverage of 74% suddenly becomes 90% but who are you fooling? So we stick with the standards and do not try and spin the numbers.

For those who do not know the dB scale as it relates to power output; Every -3dBm is half of the previous level, so going from -100 to -115 is 32-times weaker. You'll agree that's going to make a huge difference. Not 15 times, but 32 times.

The second important point revolves around the 3G phenomena of cell shrinking. This is part of WCDMA and affects all networks, even those ones running on 900MHz.

As soon as you load the cell, it gets smaller. And very quickly. Put a few users on (less than 10) and the cell shrinks to a fraction of what is was unloaded. Again, users will see this suddenly losing 3G when they used to have it. Or very low speeds. (The only way to fix this is by adding more towers, but that's another topic, although it's been discussed here recently)

To get back to radio planning, the question is then; what loading do you use to run your prediction? If you use no loading and a low signal level, you can calculate you're *best* coverage number. But you probably should not as those complaints of lost-3G and slow speeds will hit you again. And you forumites are quick to point them out!

Again, Vodacom is very conservative around this.

I respectfully ask that you edit your post (removing my name).

Below is the person who asked for proof not me.

I doubt you can prove any of your claims

No operator covers all of South Africa with 3G. But some do cover much more than others. And it's a fact that "Cell C" covers the most. "Cell C" has the most 3G towers, the biggest 21.6 network and, by far, the biggest 43.2 network. That's just a fact. And across all these network, speedtest.net shows "Cell C" has the fastest average speed. Another fact.

If you use the standard that most networks use to calculate coverage, i.e. a decent signal level that will actually give a decent connection, "Cell C" covers 74% of the population today.

If you use, like some, questionable levels to calculate coverage based on very weak signals that will probably result in disconnects, "Cell C" could claim 90%. This "90%" could then be directly compared to other doubtful coverage claims. Yet another fact.

I CAN ALSO MAKE FACTS...

 

[Mbuzi]

Man, would we (and any other operator) love to upgrade GPRS to HSPA+ via software. I see a Nobel price coming!

Why don't you post (or PM) me your location and I can check what the story is.

The 2G (i.e. GPRS/EDGE) coverage is ~99% of population today while the 3G is 74%. So there'll be plenty of areas where you'll get 2G but not 3G.

And it'll vary from network to network as not everyone have the same number of towers in the same areas or even have enough backhaul to the towers already there.

Just PMd you.