onsdag, mars 04, 2015

The spectrum Illuminati situation

In a recent blog post by the illustri(ous) Fred Wilson, mr AVC.com, he aired a bunch of opinions and questions regarding LTE in the WiFi Spectrum. Usually there are more comments to his blog posts, maybe this time it was less about having an opinion than knowing what is the actual situation that limited the numbers and made it (over)viewable. Good.

The post was mainly about T-mobiles move into the unlicensed spectrum 5GHz and the questions are:

1. Can LTE and WiFi (and everything else in the 5GHz band) play nice with each other?
2. It this works, is this a model going forward? Why not have the mobile carriers provide services in unlicensed bands versus licensing them valuable spectrum?
3. Does this provide T-mobile a cost advantage model for competing in mobile broadband (not having to spend billions to buy licensed spectrum)?

All valid questions and ideas and centric to what is going on in the radio per smartphone area, lets call it RpS for this post, rather boring to write...

So my short answer was yes, yes and no, why not and yes. But with a few more words.
There were no objections to my answer but one that said pretty much the same with fancier language.
And I said that maybe it is time to write a post about this as the f**ng Internet have nothing on the subject that concludes the core problems and solutions. If there are solutions.

I'd extend the answers with 1a) yes they can play nice if there is a new standard for how to (self) administrate the APs, Access Points and in this case I extend APs to LTE base stations because of 1b) which means that since the launch of the idea of 4G or LTE as 4G there have been theoretical and ideological discussions on how to deploy LTE and how it would look like by topographical and practical rules. In the early days the somewhat economically orientated soca experts from networks, mostly, said that "well let's just mount them on 3G masts and in context of 3G BSs, Base Stations, and the higher power from broadcasts will enable the wanted/needed speed of 100 mbits". Well, no.

Anyone with a sane view of the dysfunctional standards of 3G and still remembering the roll-out chaos said "no f**g way it will work" so a lot of money and thought was put into testings but not live testing because this industry doesn't believe in live testing until it is time or due time to deliver:- so these simple tests showed that "hmmm you might want to put a LTE BS in every street corner, in fact to be on the safe side put 4 in every street crossing on every building. And maybe 1 in between if it is a larger block."

Reminder, this is my personal interpretation and ironic view of +100 of research papers, reports and promotional press releases from prominent players, only interpreted for my own use and projects and not an industry view of the matter. End of reminder. ;-)

So a year or so before first roll-out this was the semi-paradigm. The month after roll-out it wasn't. We need more BSs! And for f-s downplay the 100 mbits idea. Say 25. Or don't even promise anything beyond 7.

Complete schuck-up.

And I'm not even mentioning the chaos on the smartphone side on how to make 2, 3 and 4 G play together with WiFi and BT.

Now today we have matured, but I'm not sure it's a good thing because the industry have settled for less. The blod ideas of what LTE could be is gone or almost gone. Now the officials are starting to talk about 5G and how that will solve the demand for almost limitless connectivity and data everywhere.
The buffoons are on the floor again.

So when LTE can't deliver 4G and spectrums cost licenses? Why not rely on technological advances and see if the engineers can figure out how to make short range radio (WiFi) work with LTE (semi-long range) and see if the cheap deployment of WiFi could help out with the humongous LTE roll-out costs?
On a technological angle this could work now, it would require a lot more advanced WiFi equpment and self-monitoring plus meshing functionality but it is doable. Wasn't that easy 2-3 years ago but now I think it is doable.
But speed... still the idea of +100 mbits speed is far far away in another galaxy.

So it is still a lousy solution to the promise.

Is it a way forward? Hell yes.

Is it economically better? 100 x hell yes!

If it gives T-mobile an advantage? In the short range yes but it's very easy to copycat it.
If WiFi is the solution for everything in the next 5 years.


No No No No No.

M2M, IoT.
There is a real problem way beyond silly consumer surf speed.

But I save my PoV on that one for now.