These are the networks which should get more 4G spectrum in South Africa

What an utterly useless mathematically cluster*****.
 
What a ****ty metric. So because Vodacom maintains high prices to keep data usage relatively low and their towers nice and empty, they deserve more spectrum than Telkom and Cell C that run these 50GB/100GB/Uncapped LTE deals that saturate their network and result in terrible speedtests?
 
This comparison doesn't make sense. The only important components are:
- amount of bandwith allocated
- amount of data transfered.

As long the "amount of data transfered" is not published, we have no any clue how efficient our networks are.
 
We added up all the IMT spectrum assigned to operators, divided the number of subscribers by the bandwidth available, and multiplied that number by a tenth of the average speedtest results on those networks.
Click to expand...
So it's all about who can give the least data to ensure the best speeds to the most people. Naturally guess who comes out on top. :rolleyes: What an utterly ridiculously useless comparison.

I've given a formula along the lines of the data served divided by spectrum assigned and challenged them to do an article on that but we get this useless one instead.
 
Only catching up on this thread now...

Ignoring factors such as number of subscribers served is folly. Ideally I would've also liked to include CAPEX, but that would be even trickier than gathering the data used here.

That said, I will try to get information on the amount of data transferred as it will make for an interesting addition to the metric. I doubt whether I will get it, and even if I do I may not get data over comparable periods.

The results, I suspect, will remain disappointing for those who dislike the big telcos. There will also be those from other telcos who argue that data volumes and speeds do not adequately measure other network quality factors such as contention and reliability.

This issue is not simple. Certainly spectrum assigned and data volumes alone don't paint a full picture, and although I would argue that SEAS provides a more nuanced perspective, it is also not perfect.

Past performance is not necessarily indicative of future performance, for instance, which is where CAPEX considerations come in. Yet the operator with the deepest pocket may not necessarily be the most effective at rolling out networks that end up getting used.

In short, for every attempt at trying to provide useful analysis on this, I can probably find a shortcoming of some kind. It's easy to come up with reasons not to do something.

Right. With the navel-gazing out the way, one last point: what this does is give some indication of where you can maximise impact. If you want to quickly improve the experience of the most cellular subscribers in the country, who do you give the spectrum to?

With that question answered we can start discussing a reasonable wholesale model to open up the sector to wider competition.
 
Jan said:
Only catching up on this thread now...

Ignoring factors such as number of subscribers served is folly. Ideally I would've also liked to include CAPEX, but that would be even trickier than gathering the data used here.

That said, I will try to get information on the amount of data transferred as it will make for an interesting addition to the metric. I doubt whether I will get it, and even if I do I may not get data over comparable periods.

The results, I suspect, will remain disappointing for those who dislike the big telcos. There will also be those from other telcos who argue that data volumes and speeds do not adequately measure other network quality factors such as contention and reliability.

This issue is not simple. Certainly spectrum assigned and data volumes alone don't paint a full picture, and although I would argue that SEAS provides a more nuanced perspective, it is also not perfect.

Past performance is not necessarily indicative of future performance, for instance, which is where CAPEX considerations come in. Yet the operator with the deepest pocket may not necessarily be the most effective at rolling out networks that end up getting used.

In short, for every attempt at trying to provide useful analysis on this, I can probably find a shortcoming of some kind. It's easy to come up with reasons not to do something.

Right. With the navel-gazing out the way, one last point: what this does is give some indication of where you can maximise impact. If you want to quickly improve the experience of the most cellular subscribers in the country, who do you give the spectrum to?

With that question answered we can start discussing a reasonable wholesale model to open up the sector to wider competition.
Click to expand...

The ones that are cheapest per GB, so Telkom and Cell C
 
Jan said:
Only catching up on this thread now...

Ignoring factors such as number of subscribers served is folly. Ideally I would've also liked to include CAPEX, but that would be even trickier than gathering the data used here.

That said, I will try to get information on the amount of data transferred as it will make for an interesting addition to the metric. I doubt whether I will get it, and even if I do I may not get data over comparable periods.

The results, I suspect, will remain disappointing for those who dislike the big telcos. There will also be those from other telcos who argue that data volumes and speeds do not adequately measure other network quality factors such as contention and reliability.

This issue is not simple. Certainly spectrum assigned and data volumes alone don't paint a full picture, and although I would argue that SEAS provides a more nuanced perspective, it is also not perfect.

Past performance is not necessarily indicative of future performance, for instance, which is where CAPEX considerations come in. Yet the operator with the deepest pocket may not necessarily be the most effective at rolling out networks that end up getting used.

In short, for every attempt at trying to provide useful analysis on this, I can probably find a shortcoming of some kind. It's easy to come up with reasons not to do something.

Right. With the navel-gazing out the way, one last point: what this does is give some indication of where you can maximise impact. If you want to quickly improve the experience of the most cellular subscribers in the country, who do you give the spectrum to?

With that question answered we can start discussing a reasonable wholesale model to open up the sector to wider competition.
Click to expand...
My suggestion stands. Two factors are most important (as per my previous post). It is how they utilise the existing spectrum resources. It is already preferential to big operators like Vodacom and MTN, as if the operator gives a wide coverage as well, it will make a MASSIVE positive effect in total data transfered. The big operators can harm themself of course by high prices, contract issues, OOB shark, unfair data expiry, roll-over policy and outdated accounting system which drives customers elsewhere (even they keep voice number).

As the above figure favour big operators, I would also introduce additional factors for comparing small operators. A total data transfered divided by the amount of spectrum and divided by total number of towers (per tower figure). This would show how efficient operator is in the areas covered by their network. We would then look for the figures how many tower were commissioned over last year (a network expasion figure) and total investment for upgrading existing equipment. If a small operator is not working on expanding their network, it is obviously not worth of currently assigned bandwith.
 
Last edited:
i call BS...

Q: What happens to demand for mobile data products in an area where there has been a fibre deployment?
A: it falls off a cliff as mobile becomes a complement rather than a substitute for fixed/nomadic solutions

Q: where is fibre being deployed?
A: in the metro areas in particular but also in a number of small towns....this year there will be a massive focus on fibre + wifi in the urban townships

Q: where is the future demand for spectrum?
A: not in the above areas where the majority of revenue currently is

...this do-or-die requirement for spectrum seems to me to have little to do with operational requirements and everything to do with excluding future competition
 
You must log in or register to reply here.
Back
Top
Sign up to the MyBroadband newsletter