FHSS Detection with Spec Analyzer
Tim Bently
identifying FHSS signals within the commercial broadcast frequency
spectrum. Please excuse a few questions.
In technical terms, how does a spectrum analyzer lock onto a signal
that constantly frequency hops?
Can this still be achieved if the frequency divisions and timing are
random?
What is the best type of spectrum analyzer for this purpose? Are any
special options needed?
Are there any technical guides available online that deal specifcally
with this type of application?
Once again, sorry for the several questions. I don't know how else to
ask.
Thank you for any advice.
Tim Bently
o...@uakron.edu
"cooperate" with the TX, unless you obtain one designed for radio
servicing that can demodulate a given application. Odds are for a
educational project at the undergraduate level you would not need a
demodulating capable unit. If you get access to a modern lab grade
one, will not drift any where near enough to cause you any issues. In
fact, for the most part, it will be more stable the then spread
spectrum transmitter is. You need to give us more data as to the
frequency band and radio service you wish to look at, as with what you
have said we could specify anything from a modified TV tuner hooked
to a ramp generator and oscilloscope to a 40,000$ instrument.. You
might wish to get on interlibrary loan a copy of the ARRL Spread
Spectrum Handbook, which is outdated but very useful from a beginning
point of view.
I'm assuming you mean at 88-108 Mhz FM, or are your looking at
Digital Radio Mondial ? Or the UK/EU service at 200-300 Mhz?
Without knowing the target application, It is impossible to suggest a
unit without knowing frequency band, desired bandwidth, and noise
floor/sensitivity issues for the instrument.
Start with this:
And This:
http://www.tek.com/Measurement/App_Notes/XYZs/
And look at home made systems here:
http://www.qsl.net/n9zia/wireless/pdf/9808035.pdf
and here:
http://homepages.wmich.edu/~cotton/SA1.html
and here:
http://www.scottyspectrumanalyzer.com/
I'm not sending you to the home-made systems to suggest you build one,
I'm sending you there to learn how a simple single stage
superheterodyne SA works...
Steve
mpm
> Timing is derived from the spec an's own clock system. It does not
> "cooperate" with the TX, unless you obtain one designed for radio
> servicing that can demodulate a given application. Odds are for a
> educational project at the undergraduate level you would not need a
> demodulating capable unit. If you get access to a modern lab grade
> one, will not drift any where near enough to cause you any issues. In
> fact, for the most part, it will be more stable the then spread
> spectrum transmitter is. You need to give us more data as to the
> frequency band and radio service you wish to look at, as with what you
> have said we could specify anything from a modified TV tuner hooked
> to a ramp generator and oscilloscope to a 40,000$ instrument.. You
> might wish to get on interlibrary loan a copy of the ARRL Spread
> Spectrum Handbook, which is outdated but very useful from a beginning
> point of view.
>
> I'm assuming you mean at 88-108 Mhz FM, or are your looking at
> Digital Radio Mondial ? Or the UK/EU service at 200-300 Mhz?
>
> Without knowing the target application, It is impossible to suggest a
> unit without knowing frequency band, desired bandwidth, and noise
> floor/sensitivity issues for the instrument.
>
> Start with this:
>
>
> And This:
>
> http://www.tek.com/Measurement/App_Notes/XYZs/
>
> And look at home made systems here:
>
> http://www.qsl.net/n9zia/wireless/pdf/9808035.pdf
>
> and here:
>
> http://homepages.wmich.edu/~cotton/SA1.html
>
> and here:
>
> http://www.scottyspectrumanalyzer.com/
>
> I'm not sending you to the home-made systems to suggest you build one,
> I'm sending you there to learn how a simple single stage
> superheterodyne SA works...
>
> Steve
I'm not aware of any FHSS in the Commerical Broadcast Band (US: 88-108
MHz) ?
Except possibly for very low power Part-15 emissions.
-mpm
Tim Bently
>Without knowing the target application, It is impossible to suggest a
>unit without knowing frequency band, desired bandwidth, and noise
>floor/sensitivity issues for the instrument.
>
Thank you for your informative reply. There is no particular band of
interest, as such.
The idea was simply to "sniff" around the radio and IF broadcast bands
and demonstrate objectively how it is utilized by FHSS.
IOW to find FHSS "hidden" information content and folllow it as it
shifts from carrier to carrier, noting the spread, time intervals,
etc.
If the modulating signal could also be demodulated/isolated, that
would be a plus. Now that I think of it, I am wondering to what extent
this is actually possible given privacy issues with telecoms.
Of course, it would be desirable to demonstrate this over as wide a
bandwidth as possible, but without making the exercise inordinately
difficult.
Any further suggestions along these lines would be much appreciated.
Tim Bently
Mark
without knowledge of the hopping, it is very difficult for a standard
swept spectrum analyzer to follow a FHSS signal, finding this out
first hand may be the point of the exercise that you were assigned.
There are analyzers called "real time spectrum analyzers" that
basically record a range of frequencies and can analyze them after the
fact. With one of these you can start to analyze an unknown FHSS
signal.
Mark
o...@uakron.edu
redistribute or discuss the demodulated content.
Since you do not have a target system in mind, at this time I would
strongly suggest a copy of the ARRL spread spectrum handbook, and
build some of the experiments in it, as most of the parts are still
available and some of the spread spectrum sources detectors can be
built for less then 50$.
IFs for most FH/DS radios are 70 to 200 Mhz. If you look at the
stats for the various IEEE 802.11 standards you can get a idea of
the bandwidths involved. A issue might be that modern radios might
have the IF contained on chip, you might need to find some older ebay
stuff, for example early Telxon data links, to find discrete IFs, or
chipsets that use external SAW filters.
If all your looking for is the carrier spikes, a spec an with a
minimum bandwidth of 10-20 kilohertz is probably just fine and you
probably could get away with 100 khz.
Steve
o...@uakron.edu
http://www.sss-mag.com/ss.html
You might want to see what you cna find on the old Harris PRISM
chipset
Steve
Tim Wescott
> As part of a university course we are interested in monitoring and
> identifying FHSS signals within the commercial broadcast frequency
> spectrum. Please excuse a few questions.
>
> In technical terms, how does a spectrum analyzer lock onto a signal
> that constantly frequency hops?
It doesn't. Is that too technical?
> Can this still be achieved if the frequency divisions and timing are
> random?
It can't be achieved at all with a spectrum analyzer.
> What is the best type of spectrum analyzer for this purpose? Are any
> special options needed?
_If_ there is a plug-in or option for the spectrum analyzer that is
_specifically_ designed for the _specific_ spread spectrum service in
question then that would certainly make it possible. But then the
instrument would be a lot more than a "Spectrum Analyzer", and it's
operation would be explained in it's operating manual.
> Are there any technical guides available online that deal specifcally
> with this type of application?
Dunno -- you're asking a spectrum analyzer to be a very different
critter than it is now.
> Once again, sorry for the several questions. I don't know how else to
> ask.
>
> Thank you for any advice.
You need to build or find a receiver for the specific spread spectrum
service that you're trying to monitor. Then you need to decide what
'spectrum' you want to measure -- the spectrum as emitted from the
transmitter? The spectrum after despreading? What?
--
Tim Wescott
Control system and signal processing consulting
www.wescottdesign.com
o...@uakron.edu
the spect an display. You just cant demodulate it, and only a crude
analysis can be performed. Which is what I suspect your instructor is
trying to teach you.
Steve
Tim Wescott
(Context please, this is a USENET newsgroup, even if you're using Google).
Possibly. A spread spectrum signal, properly done, could be secreted
inside the 'normal' FM band and wouldn't even show up above all the
'real' FM activity, yet would still show up just fine after being
correctly despread and demodulated.
You'd have no chance of seeing it with an antenna stuck onto the
analyzer -- you'd only be able to see it if you could tap straight off
the transmitter.
o...@uakron.edu
Sorta, depends on the method of spreading and the bandwidth. Some you
can see quite well in the near field. Or you'll see the Barker code at
startup and sync. Depends very much on the system.
Steve
Tim Wescott
Well, I sorta agree with your disagreement, but only a bit.
A spread spectrum system that was _trying_ to hide in the FM broadcast
band could probably do so quite effectively. A system that is meant to
be public but happens to have qualities that would hide it (E.g. CDMA
without explicit despreading) would also hide effectively, at least from
a spectrum analyzer (and yes, CDMA isn't frequency hopping -- but still).
Granted, you'd be able to see its spectrum if you're in the near field
-- but that's getting close to my 'hook up to the transmitter' case.
Joel Koltner
"Tim Bently" <timb...@peromax.com> wrote in message
news:4ba56f04...@news.tpg.com.au...
> In technical terms, how does a spectrum analyzer lock onto a signal
> that constantly frequency hops?
(As Tim mentions, it doesn't -- at their cores, spectrum analyzers just
display "what's there," it's only with additional software that they try to be
"smart" after follow hops or whatever.)
> What is the best type of spectrum analyzer for this purpose? Are any
> special options needed?
Tradiational spectrum analyzers literally sweep in frequency, displaying
incoming signal power vs. frequency as they go. There are some tricks you can
use, playing with sweep rates and maximum (input) hold features to let you at
least see the various frequencies in use -- or at least the frequency ranges
in use. However, the better approach (for spread spectrum) to really see
what's going on quickly is to digitize an entire chunk of spectrum over a
relatively brief interval of time (on the order of a small part of the hop
time) and then display the FFTed result on, e.g., a waterfall display. If
that "chunk of spectrum" is small (tens of kHz or less), this is easily done
with pretty much any radio receiver and a PC (you feed its audio output to the
PC's sound card) -- this is often called a "panadapter." If the chunk of
spectrum is large (which is usually the case -- "spread spectrum" being about
"spreading out" the spectrum and all :-) ), you need high bandwidth receivers.
Hence you need a somewhat more specialized receiver ... or if you're still
after a spectrum analyzer as such, a so-called "real time" spectrum analyzer
like these: http://www.tek.com/products/spectrum_analyzers/
(You can find much cheaper units avaialble if you have specific uses in mind,
such as just looking at Bluetooth -- one of the most common FHSS transmitters
people tend to own.)
> Are there any technical guides available online that deal specifcally
> with this type of application?
What are you trying to do? Tek's web site there certainly has various guides
that might help...
---Joel
artie
<timb...@peromax.com> wrote:
What's the center of the band, and how wide is the band? You'll need a
wideband receiver with some interesting characteristics.
Google Agilent's Blackbird -- mostly sold to Three Letter Agencies,
very good at finding hoppers.