[LINK] Tech startups release manifesto for policy change after government's Policy Hack

Paul Brooks pbrooks-link at layer10.com.au
Mon Nov 16 15:06:26 AEDT 2015

On 16/11/2015 9:41 AM, Tom Worthington wrote:
> On 15/11/15 16:33, Paul Brooks wrote:
>> All 'mobile' infrastructure has a foundation of fixed infrastructure. ...
> Yes, *almost* all mobile devices require a fixed infrastructure. But building a
> fixed infrastructure all the way into people's homes is not an efficient way to
> support their mobile devices.
No - all mobile devices need one or more networks that enable the device to
communicate while mobile.

Efficiency is a completely separate beast and depends entirely on the metric - a
network that is efficient at providing connectivity while the device is stationary at
ground level may be completely unfit and inefficient at providing connectivity while
the device is travelling at 100km/hr, or 3000m below the surface of the ocean, or
10,000 metres above the ground, or while located in a different location - and vice
versa. A network efficient at providing kilobit-per-sec capability over hundreds of
kilometres may not be efficient at providing gigabit-per-sec capability over hundreds
of metres. There is no such thing as 'one size fits all use cases' for network
infrastructure, whether fixed or mobile. Efficiency is also not related to
effectiveness - a network that is efficient for communication may not be able to do so
effectively everywhere, regardless of how efficient it might be.

>> ... NBN fixed infrastructure forms a fine backhaul network to enable
>> ubiquitous WiFi and other radio technologies to connect mobile
>> devices to educational content and to each other. ...
> Most of the cost of the NBN is in the last few hundred metres, getting fiber from
> nodes in the street into homes. If this part was wireless, it would eliminate a
> major cost.
Thats a disputable opinion Tom - to be a factual statement it needs to have context
and qualification.

If this was wireless, it would be ineffective and not fit for purpose for many
people's purposes, and you should choose a different metric. If this part was
wireless, and provided the same utility and capability as the fibre component (or even
copper, since we're discussing fixed line), it would cost more than the fixed
component, not less.

> The idea that you use one fixed data network at home and then you to switch over to
> a different "mobile" one when you step out the front gate seems an antiquated idea.
> Hardly anyone does that for making phone calls any more, so why should should they
> do it with data?

Because the use-cases and destinations for data transfer are completely different than
for making phone calls. "Phone calls" are a particularly bad exemplar, because they
don't have any use-case for communicating with another phone located in the same room
or home.

The notion that traffic between my mobile devices, storage systems, display devices,
home sensors and the exercise monitor on my wrist should be forced to hair-pin on low
bandwidth, lossy and variable links across a suburb or across a city and back again
when they are physically located a couple of metres away from each other is ludicrous
and inefficient.

The idea that a mobile device can only use one type of network seems an antiquated
idea. Mobile devices contain many interfaces and can use many networks, and are
generally smart enough to use the network that is best suited for the task at hand,
depending on the location and demand. I *want* my mobile device to switch from the
low-capacity suburb-scale radio network when I'm out of my house to the high-capacity
house-scale network when I am inside it, so the data can travel freely, securely and
without tarriffs between my device and my other devices located close by, within my
security perimeter.


More information about the Link mailing list