Re: 5G roadblock: labor

[Mark Tinka <mark.tinka () seacom mu>]

On 6/Jan/20 23:46, Andrey Kostin wrote:

> I'm talking only about last mile access.

As a last mile technology, yes, wireless is fine. We use it today for
4G/LTE; it is a last mile.

But as a backhaul technology, it won't do. You need wire for that, at
least in 2020 anyway.


> Wireless is going the same path as fixed access before: from big
> central facilities to end-user as much close as provided services
> bring enough revenue to cover upgrade costs and create some profit.
> With copper phone lines the situation has already turned backward
> because revenue from services isn't sufficient.

In Africa, most homes use some kind of 3G/4G/LTE router to get broadband
into their homes. This mainly due to a lack of fibre in one's specific
area. But every time fibre shows, up, they switch over, purely because
the performance of the GSM network is unpredictable, and the mobile data
costs are too high for a modern home in 2020.


> We all know physics and Shennon/Kotelnikov theorema. To get more speed
> more spectrum is needed but more spectrum is available in higher
> frequencies, that have shorter coverage.
> Where it's going to stop - I don't know, 6G or 7G or XG ;) Only making
> enough money is needed to go to the next G.

And the cost that goes along with it.


> Regarding comparison WiFi and cellular networks, it's clear that WiFi
> won't be able to compete with mobile in terms of scalability.

We aren't talking about trying to drive an entire nation on wi-fi with
towers. 5G or any other G operating at higher frequencies suffers the
same constraints.

We are talking about using wi-fi for dense locations where 5G also makes
sense technically, but not commercially, yet.


> Building WiFi in public places like stadiums is already became a job
> specialization, but every such implementation has it's limit. On the
> other hand, 5G as I can see is a big step in this direction in terms
> of spectrum and subscribers management. Mobile networks are developed
> for central control of all the components on all layers, that's why
> mobile standards contain thousands pages. WiFi is a technology for
> local access and to make it more scalable means to go through the same
> development process as mobile networks did. Something can probably be
> improved but even if it succeed it won't be cheap anymore. Currently
> WiFi is only describes single layer of connectivity, and this is why
> it's cheap, but on the next layer (i.e. IPv6 implementation) we can
> see incompatibility between "standardised" WiFi devices. Compatibility
> on many layers is necessary to orchestrate all of them, so not going
> to happen. Yes, WiFi and mobile can be compared in radio, but not in
> anything else.

There are some wi-fi vendors who I know (and am currently testing) that
have developed very cool centralized management tools for their wi-fi
AP's, that include very interesting AI logic. It is pricier than a
simple standalone enterprise-grade AP, or an AP you'll get from down the
store. But it's still way cheaper than dense 5G deployment.

Mark.