Local Only 3G / HSDPA Access

Ive downloaded a couple of 100MB or so using Vodacom HSDPA over the SAIX News Server.. normal internet APN.
 
A news server is definitely the way to go, but then we need cheap local access. I understand why we cannot get free local access, but a cheaper option will be better as the prices are still way too high. V3G, when is Voda's next drop in data prices? I'm on a 2GB bundle @ R449pm and let me tell you its not nearly enough and I don't even download huge files. The only stuff I download are mp3s from iTunes. I really applaud Vodacom for the service they are providing, it is excellent. I'm in an Estate where Telkom is not available, but thank God for HSDPA.
 
MTN restricts the uplink speed on its network to 64kbits/s, whereas Vodacom allows for the full 384kbits/s HSDPA uplink speed, so obviously Vodacom is different.:cool:

My bundle on MTN is almost finished, so I'll soon be trying Vodacom prepaid, hopefully its better.

A little O/T: Can someone tell me how to by a prepaid Vodacom data bundle? I obviously need a sim card, but I havent had a prepaid anything for about 10 years.
 
:confused: in English?

Yea, It was almost 2AM. No wonder I made no sense. You need to remove the "month-end"/"data-bundle-end" rush of data usage from the stats you get. In order to get the realistic stats, People using the saix news server just to use up their bundles, cannot count part of the final stats.

Is that more clear?

A little O/T: Can someone tell me how to by a prepaid Vodacom data bundle? I obviously need a sim card, but I havent had a prepaid anything for about 10 years.

http://mybroadband.co.za/vb/showthread.php?p=1086462#post1086462
 
It would be great from a business point of view if we could get a local only APN type thing... I have at least 20 data cards where they have been bought for local VPN usage but given the cost involved they have to be restricted to 500meg.

Trying to think of useful ways to implement an off-peak or some type of local system where we can get more bang for our buck by keeping things on Vodacoms network, or at the very least staying inside SA's network.
 
Just a question: Is it your international transit or your transmission and access network that you are trying to optomise during off peak times?

Both, I would think. As ic pointed out the cost of the lines are constant, irrespective of loading, so an underutilised network is a massive waste.

So it's not much optimising it, more the opportunity to rev it 24/7....

Thus the question, how do we use the slack time to everyones advantage?
 
Ok, I understand what you're saying, but I don't think that's feasible for several reasons, in particular staggered expiration of PPDBs and there being more PPDB customers than there are CDB customers.

Pulling the stats of the CDB customers and removing the last +-3 days, should be a joke. Doing the same with the PPDB customers, could be a bit more tricky, but if the system was designed to be open and modular (designed correctly iow), there should be no problem doing this at all. Could maybe just take a tad longer than pulling the CDB's stats.

[edit]
But then again, I'm not VC - I dont have a clue what their systems look like.
 
Last edited:
Thus the question, how do we use the slack time to everyones advantage?
I think that ic's suggestion of discounted off-peak rates is a winner. It provides cheaper bandwidth to all of those who are so desperate for it and additionally it provides a real incentive for those running automated download processes to move their traffic out of peak times. If I need my machine to download a massive chunk of data once every 24 hours and I don't care how quickly that data is coming in then there is no reason for me to choose an off-peak time to do it - despite the fact that I am affecting the network performance negatively for those interested in real time applications.

I think that the idea of mirrors/caches/news servers is an extremely good one but one that must be carefully considered. TENET already has an excellent mirror server that they update using their off-peak international capacity. I think however that its true value is underutilised and to setup another mirror just to 'use up' Vodacom's surplus capacity would be counterproductive and dilute the effectiveness. Same goes for a news server - it should be peered with SAIX's server (and perhaps even IS's as well) and then only download what is not available locally. Get you much more bang-for-your-buck that way too!

That said - I think a Vodacom news server with a low rate off-peak pricing structure would be well received. Additionally - I think that the upstream (international) bandwidth for the server could be controlled to correspond with the off-peak periods on the VC network. So if an item is cached on the server then it goes out at full speed but if it needs to be pulled from upstream then it is explicitly set to be faster during the off-peak periods. More incentive for off-peak usage :D
 
Last edited:
Was also smiling at that one :)

We try and do a lot for local broadband (might not always be obvious) but giving the family jewels away is asking a bit much.

@ic, working on the Vodacom site issue. Nearly there...;)

As always, my question was a leading one. So come up with good ideas and let's discuss. I'm still keen to do something with off-peak bandwidth....

MTN have/had a system where on-net calls were discounted by a certain amount depending on current usage etc, couldn't something similar work for data?
 
On-net voice calls don't carry the interconnect charge, so tend to be less expensive.

I know, what I meant is that they varied the discount offered based on current network load to encourage people to use the network during off peak times, couldn't the same principle be applied to data usage with adjustments? From what I understand both on-net voice calls and data usage generally would have a similar business model as the network costs for a given capacity/speed are fixed regardless of actual utilisation, unless I'm missing something?
 
I know, what I meant is that they varied the discount offered based on current network load to encourage people to use the network during off peak times, couldn't the same principle be applied to data usage with adjustments? From what I understand both on-net voice calls and data usage generally would have a similar business model as the network costs for a given capacity/speed are fixed regardless of actual utilisation, unless I'm missing something?

Agree, just pointed out that with voice there are additional reasons for different on-net / off-net charges.

With data, this is not the case, so the model will need to accomodate it. (or the lack of it, rather).

Think we're all saying the same. ;)
 
Agree, just pointed out that with voice there are additional reasons for different on-net / off-net charges.

With data, this is not the case, so the model will need to accomodate it. (or the lack of it, rather).

Think we're all saying the same. ;)

Think so. :) I just thought the idea of a dynamically changing tariff based on current usage was an interesting idea rather than the standard peak/off peak pricing you usually see with voice calls. Its probably harder to implement and more confusing for consumers though.
 
Its probably harder to implement and more confusing for consumers though.
Do you recall the three tier tariff that telkom used to have many years ago? Caused huge amounts of confusion as I recall.

I would personally think that the simplest possible tariff plan would be the best.
 
Do you recall the three tier tariff that telkom used to have many years ago? Caused huge amounts of confusion as I recall.

I would personally think that the simplest possible tariff plan would be the best.

Was that with the distance bands? I agree but in the MTN model you had the standard peak and off peak tariffs but then received a varying discount on those, so it was in effect an extra over and above the standard rate. It was easier to communicate that though as it could be displayed on the cellphone screen, there's no similar functionality on data cards and data sessions also tend to be longer than voice calls so the model might not transfer very well.
 
So, how long till we can expect any changes in local MB tarrif?
 
So when is Vodacom going to let us surf local for free? I think Vodacoms' high prices should at least include the free local bandwidth. Come on Voda, dont tell us you have no means of splitting International and Local, thats a feeble excuse. Get off your behinds and get us free local bandwidth, we are paying you a fortune.:(
Local will only be free once all internet is free, i.e. uncapped for a fixed price. Personally though I would prefer a price reduction and since it is now 8 months since the last one and the user base is a lot larger it is due for another reduction. Considering Neotel will be entering the wireless market it would have to be soon depending on what they offer. I would say 5GB for R300 which is more inline with the international market Vodacom likes mentioning.
Some examples where local browsing is desireable?
Connection to a company server or website perhaps. The internetvpn apn can be revamped for this purpose with the internet apn for general browsing and the unrestricted apn for more demanding users. On network access could be even cheaper with no peering links.

Seeing you agree it's easy enough (your words) to split the two a package where 10% is for international should suffice. Local would not count towards the international cap but international would count towards both.
Both, I would think. As ic pointed out the cost of the lines are constant, irrespective of loading, so an underutilised network is a massive waste.

So it's not much optimising it, more the opportunity to rev it 24/7....

Thus the question, how do we use the slack time to everyones advantage?
Something some of us have suggested in the past. Since the bandwidth is paid for whether used or not (depending on peering agreements) might as well provide an incentive to use it. Something like charging 1/2 when usage is 1/2 or 2/5 when usage is 2/5. It would be helpful to know more about the network particularly how the international and local components works. IS doesn't keep this a secret so I see no reason for Vodacom to do so.
 
although not every part of the local net is reachable via JINX & CINX, but it would be a very good start.
Incidentally - CINX is not currently active. However you touch on an important point: It is fairly easy for the national first tier operators (SAIX and IS primarilly) to provide cheaper local access that runs over their own network and peering links. The '1,5' tier and second tier operators will find this a lot more difficult as they often have to purchase upstream capacity (more expensive) in order to make sure that they are getting access to all the local providers that they don't peer with. This is also likely to be further complicated by the increase in the number of first tier operators with the new ECNS licensing framework.

We are going to end up with "local only" products that are either not connected to everything locally or are more expensive.

Connection to a company server or website perhaps. The internetvpn apn can be revamped for this purpose with the internet apn for general browsing and the unrestricted apn for more demanding users. On network access could be even cheaper with no peering links.
I think that an on net APN would be a useful addition to the Vodacom data products. Its the nearest to local-only that VC could guarantee with their current network peering. It would also provide a very useful intermediate product for companies that don't qualify for corporate APN services. And once they get some fiber in the ground they could provide fiber access into this cloud. This would provide a great, cost effective VPN solution for small and mid-sized companies.

Seeing you agree it's easy enough (your words) to split the two a package where 10% is for international should suffice. Local would not count towards the international cap but international would count towards both.
Provisioning 10% of your capacity for international is something very different from not charging for local. There are completely different parts of the network infrastructure responsible for routing and billing :cool:
 
As always one needs to be very careful when building these solutions. Every single time we think we've thought of all the angles, someone comes up with one we did not consider, or gave a low priority. Thus the long cycles for new products. What seems simple as a conceptual idea, can, and often does, become vey complex.

For example, it would be the easiest to have two APN's, but how many people know how to switch. What will happen if you think you're on the cheap APN and download a sh!thouse of data, only to discover you never switched?

But we're busy looking into this at the moment.
 
Top
Sign up to the MyBroadband newsletter
X