Re: [LINK] 'Parasitic grid' could undermine wireless revenues

From: Glen Turner (glen.turner@aarnet.edu.au)
Date: Wed Aug 29 2001 - 10:43:34 EST


Michael Skeggs wrote:
>
> Isn't the idea of this spread spectrum wireless that it does not need to be
> licensed? Or is the fine in regard to acting as a telco without a licence?

Hi Michael,

There are two sets of legislation involved. All this is from
memory, so the detail may not be 100% accurate.

The Radiocommunications Act requires transmitters to be licensed,
and some operators of transmitters. For 802.11 the transmitters
have a "class license" that covers the entire category of 802.11
devices with no per-device fees. The users don't need a Certificate
of Proficiency, but equally the users aren't permitted to alter
the operating characteristics of the device (like changing the
frequency or power).

The Telecommunications Act doesn't allow people to offer transit
to people outside of their 'immediate circle'. This applies
to any electromagetically-based communications -- wireline,
fiber, wireless and infra-red.

It the Telecommunications Act that 'parasitic grid' suggestions
run into. Essentially the operators of these devices must
become a carrier before offering transit services.

The UK legislation operates in the same way. The NTK article
points out that restricting transit services to carriers is not
socially desirable. NTK didn't choose to mention the upside
of telco-oriented communications -- mainly a large investment
in engineering leading to robust infrastructure.

Although the 'parasitic grid' model is attractive it has some
downsides:

  1) 802.11 isn't up to it. It has absolutely zero intelligence
     about distributing traffic across cells, so the system will
     fail once loaded. We've found that when deploying 802.11
     at our conferences that we need to design the 802.11 network
     and then ban everyone else from bringing base stations. That
     doesn't bode well for an organic model of 802.11 networking.

  2) The reliability won't approach that of carriers. The only
     deployable protocol is 802.1D spanning tree, and as a result
     each failure will cause an outage of at least 15 seconds.

  3) What is the routing design? Most 802.11 networks currently
     operate as a single subnet to allow easy roaming across
     basestations. There is nothing in the design for handover
     at cell/subnet boundaries, but putting all base stations
     in one big subnet won't scale.

Don't get me wrong, I think that 802.11 is going to be huge.
And there's every chance of it making G3 redundant before G3
is even deployed; much as we've seen with OSI v. IP or
ATM v. ethernet.

But there's a lot of work yet before 802.11 is developed to
the stage where it can be used as a basestation in an organic
network.

-- 
 Glen Turner                                 Network Engineer
 (08) 8303 3936      Australian Academic and Research Network
 glen.turner@aarnet.edu.au          http://www.aarnet.edu.au/
--
 The revolution will not be televised, it will be digitised



This archive was generated by hypermail 2.1.1 : Fri Aug 31 2001 - 03:10:05 EST