Re: [Discuss-gnuradio] AOR JT2000 wide band DSP FFT receiver

[Dave Emery]

On Wed, Jul 25, 2001 at 08:30:16AM -0700, Steve Schear wrote:
> >
> >Price £ T.B.A. but primarily targeted at the commercial and government 
> >market, possibly around GBP 2.5k
> 
> Looks like a good target for comparison.  By removing the case, power 
> supply and other "unnecessary" items a GnuRadio with comparable performance 
> could be much cheaper.  In my back-of-the-envelope thinking recurrent 
> manufacture costs, in small volume (100s/year), of a much less than $500 
> year (including broadband preselector and downconverter and high-speed 
> signal processing board).
> 

        That is consistant with my guess as well -cf the Winradio
boards. I suspect the markup (gross margin) on low volume specialized
products like these by AOR and others is very substantial - it almost
has to be if they sell 100s to low thousands and spend considerable
dollars developing the product - probably especially on software and
firmware (how many developers at $100+K a year can you support from
such a low sales product, and how much can you expect each developer
to accomplish ?)   Just advertising, marketing (trade shows etc), and
distribution costs  add a lot if your sales are small - even agency
approvals (FCC etc) aren't cheap.

        I am sure that a hypothetical gnuradio board could be built for
a manufacturing cost on the order of your $500 in small (100s)
quantities if most of the design labor and all the software and firmware
was open source contributed labor, though I haven't done a lot of
pricing of the various chip sets involved (and low volume chip prices
are a political thing for manufacturers, sometimes they want big markups
over volume prices to cover sales costs).   Of course nobody would be
making any money off the product at that price, and an awful lot of its
labor content would be contributed for free by those who want to see
such a thing exist and available (eg rf hackers).

        This of course raises two issues - there is a small market of
the brilliant and sophisticated who can both contribute to the
intellectual or physical capital of the project and are interested in
hacking with software radios, and a much much larger market of scanner
listeners, hams, people setting up rf freeband networks and the like who
might purchase a prepackaged "script kiddie" type solution if it was
available and required little technical skill on their part to use.

        Potential sales to the first group are very small - essentially
a prototype type business - potential sales to the second group are much
larger, especially if the devices allows such currently unavailable 
access as APCO 25 digital voice on police networks and various kinds of
digital wireless systems.   The problem is, of course, that many of the
most interesting applications to this second group of customers are
either blatently illegal in the US (digital cordless and cellphone
interception) or protected by patents and other IP (APCO 25 scanning).  
Thus any attempt to offer the board to the second "user" group along
with firmware allowing any kind of interesting access is likely to run
smack into serious legal problems both criminal and civil. 

        And the current FCC rules on scanning receivers sold to the
general public make for some very interesting issues in a broadband I
and Q  downconverter type design.  If access to the chips and firmware
is  provided as part of the package it seems impossible to meet the FCC
requirements that it be essentially impossible for an end user to modify
the device to intercept cell calls without destroying it.  And silly as
this regulation is, it is the law in the USA and enforced.   Selling a
board that provides a general purpose DSP receiver platform to the
general public seems like something that is essentially legally
impossible if it covers either 800 mhz or 1900 mhz or can be modified to
do so by someone with normal technician tools and skills (and software
and/or firmware from the net which he clearly does not need to be
able to create or even understand). 

        It is of course possible to sell such devices to researchers and
developers as test equipment and people in government and the
communications industry for monitoring and it is legal to sell ham radio
transceivers without requiring that the frequency control components be
sealed and unmodifiable  - but only to hams as ham radios.  Anything
else requires FCC type approval and comes with an explicit prohibition on
user modifications that cause the device to fail to conform with the
rules, especially regarding cell coverage. It is unclear to me whether
any form of software defined radio in which the software is accessible
to the public will ever pass FCC type approval. 

        I fully expect that AOR will never be allowed to market their
radio to the public in the USA - I bet they don't even intend to try.
They know that most of the seriously interested people will illegally
import them from Europe or Canada, and that at the price most of their
market is spooks and the military for surveillance and perhaps 
communications carriers and manufacturers for various specialized test
equipment applications anyway.   And if they do ever try to sell them to
the public, there will certainly have to be some kind of seal put on
general access to the frequency control chips and quite possibly to
downloading your own firmware at all - and this will have to be
demonstrably strong enough so it presents a serious barrier to users (at
least in the mind of FCC officials).

        Even selling kits of parts has run into FCC and DOJ problems.

        Thus I am afraid that any gnuradio board will either have to be
built and sold outside the USA and not marketed to the  US public (and 
of course if it becomes popular with US customers it may well become
considered contraband which can be searched for and seized by customs
and the FCC and for which the FCC can levy substantial ($10K or more)
fines).  Or it will have to have a transmit capability and be rather
carefully marketed and sold only to licensed amateurs for use on ham
bands with a great deal of effort made to prevent anyone from advertising
or being very public about offering firmware that allows it to be
used as a scanning receiver or even a part 15 unlicensed intentional
radiating device (both of which require type approval).

        The other issue of course is what the actual designers and
developers feel about allowing someone (perhaps not one of them) to
market and sell their technology and make money at it.  Thus if anyone
finds a way around the legal issues and taps the second "user" market
there will be a point when the limits of the GPL as it applies to hardware
and firmware and other intellectual property will doubtless be tested.
Of course for GNU open source developers, this is a rallying cry and
a political cause, so I don't expect most participants here to see this
as a problem.

        Of course I am a certified pessimist, and I hope I am wrong
and misguided about this... I don't thing I am as far as the FCC rules
go, however.  One can hope that some enlightend soul at the FCC will
see what this does to the kind of private entreprenurial experimentation
that has built our technology base and relent, but I wouldn't count on
it with the CTIA breathing down their neck.


-- 
        Dave Emery N1PRE,  address@hidden  DIE Consulting, Weston, Mass. 
PGP fingerprint = 2047/4D7B08D1 DE 6E E1 CC 1F 1D 96 E2  5D 27 BD B0 24 88 C3 18