FHSS Detection with Spec Analyzer

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 You might wish to get on interlibrary loan a copy of the ARRL Spread Spectrum Handbook, which is outdated but very useful from a beginning point of view.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:

And look at home made systems here:

and here:

and here:

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

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

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

NO issues with demodulation in a lab in the states if you do NOT 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

here, This is outdated, but it is a good start

You might want to see what you cna find on the old Harris PRISM chipset

Steve

It doesn't. Is that too technical?

It can't be achieved at all with a spectrum analyzer.

_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.

Dunno -- you're asking a spectrum analyzer to be a very different critter than it is now.

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?

For the record, in the near field, you can see a FH or DS signal, on 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

(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.

Tim, respectfully must disagree a bit.

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

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.

(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.)

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:

(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.)

What are you trying to do? Tek's web site there certainly has various guides that might help...

---Joel

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.