Spread Spectrum?

Look up UWB ("Ultra wideband").

Yep. Every Wi-Fi chip made since about 2001 has been totally digital.

Now, if you want "truly wideband" spread spectrum, there is UWB (Ultra WideBand): which goes from 3.1 to 10.6 GHz. Is that "truly wideband" enough?

More like day dreaming. Get thee back to work and do something useful, even if it's wrong.

--
Jeff Liebermann     jeffl@cruzio.com 
150 Felker St #D    http://www.LearnByDestroying.com 
Santa Cruz CA 95060 http://802.11junk.com 
Skype: JeffLiebermann     AE6KS    831-336-2558

All done with ECL...

...Jim Thompson

--
| James E.Thompson                                 |    mens     | 
| Analog Innovations                               |     et      | 
| Analog/Mixed-Signal ASIC's and Discrete Systems  |    manus    | 
| San Tan Valley, AZ 85142   Skype: Contacts Only  |             | 
| Voice:(480)460-2350  Fax: Available upon request |  Brass Rat  | 
| E-mail Icon at http://www.analog-innovations.com |    1962     | 
              
I love to cook with wine.     Sometimes I even put it in the food.

--

Jeff, I liked your poem on your website about hackers:

"I am the hacker. I break all the rules. I'll tear up your system. I have all the tools. No system is safe, when I'm on the phone. I do it for pleasure, and all on my own. I've been in your files, and made the grand tour. I've stolen your secrets, of this I am sure, and when I'm discovered, they all act like fools. for I am the hacker. I now make the rules."

Jeff, about day-dreaming, do you know what "projection" is?

:)

Of course. No need to read between the lines. My poetry should tell you something about me. Incidentally, I only write poetry when I'm frustrated, irritated, bored, disgusted, confused, etc. and I haven't written much for many years. However, I'm now attempting to slog through my income taxes, and am seriously tempted to start writing poetry again. Where are the diversions when I need them?

--
Jeff Liebermann     jeffl@cruzio.com 
150 Felker St #D    http://www.LearnByDestroying.com 
Santa Cruz CA 95060 http://802.11junk.com 
Skype: JeffLiebermann     AE6KS    831-336-2558

I wonder if the OP meant DSSS versus OFDM?

Let me clarify. I was interested in the extreme case, where the xmitter is xmitting at frequencies from 100 khz to 10 ghz. (just example numbers, nothing more.) Then the receiving is done via the digital sequence.

The original conception of SS, as I understood it, was that the large expanse of frequency made the signal indistinguishable from noise in any narrow band width. In that respect, the frequency nature of the signal is more similar to what a car ignition or electric motor puts out.

So my question was if anyone has experimented with that full expression of SS, obeying all laws of course.

From memory, Bob Friday of Ricochet fame, once declared he was very happy they had opted for [frequency hopping?? forgot the name because the names seem reversed, but it's where you simply jump the carrier around over a given bandwidth], instead of gently splattering the energy over the bandwidth. If that's any help.

I've seen both used while measuring ambient noise around the Bay Area, but most of the commercial SS systems which usually come in at lowest costs appear to all be freq hoppers and the more expensive systems, those used by utilities companies and military appear to be splatter type [is that called direct spectrum?] anyway, on a spectrum analyzer the hoppers look like that auto engine noise and the 'gentle' ones look like a 'lump' in the noise floor. To me the hoppers are noticeable, the 'lumps' no quite so noticeable.

My guess is that it's fairly difficult to SS over more than an octave, maybe a decade at most, only because of the logistics of the antenna design. It's not that easy to match 50 ohms efficiently to 377 ohms over

100kHz to 10GHz. The antennas I've seen usually come in loops, rod, multi-element, and then horn at the top. Difficult to get one to do all. Probably better to simply tag along with the microwave oven AC mains bursts and insert a small not noticeable 'packet' of info. like a very, veyr clever bluetooth synced to the AC mains.

yes. look up low probability of intercept

Metricom used FHSS (frequency hopping spread spectrum) while everyone else used DSSS (direct sequence spread spectrum). Each had their advantages. The problem of which technology to use was neatly solved by the FCC, which apparently decided that DSSS was best and sealed the victory with a collection of rules that limited the ability of FHSS vendors (Agere, Proxim, Raytheon, Raylink) to compete effectively.

I'm one of the die hards that believe that FHSS Wi-Fi is generally superior. Try to jam FHSS and it gets through (although slowly). Do the same with DSSS, and it rolls over and dies. When FHSS meets DSSS, FHSS always wins. DSSS is easily sniffed at the MAC layer, while FHSS, with it's multiple code sequences allows multiple simultaneous "channels" without much mutual interference.

Please not that Bluetooth is FHSS. While FHSS is required to hop over

79 of the 80 available channels before repeating the PN sequence, DSSS can operate over about 1/3 of the band (thanks to the FCC rules). The error was recognized by the FCC and Bluetooth can also hop over part of the band.

Another reason for the fall of FHSS is lack of speed. While

54Mbit/sec association speeds (about half that for throughput) are common with DSSS, FHSS never really made it past about 3Mbits/sec. That's not to say that it couldn't be done, it just wasn't done. Meanwhile, things have changed radically in Wi-Fi. The emphasis seems to be on speed, which nobody really needs or can use. That usually means occupying as much bandwidth as possible and maximizing the bits per baud. Some systems even use both the 2.4 and 5.6GHz bands simultaneously to improve speed.

I used to drag a spectrum analyzer up to Loma or Umunhum. The 83.5Mhz of the 2.4Ghz band was solid junk, even 15 years ago. It's probably worse today.

FHSS looks like a mess of hash, all the same amplitude over the entire band. DSSS, looks somewhat like sin(x)/x or:

I'm rather surprised that you were able to recognize individual transmitters among the junk floating around the Bay Area. To me, it was like trying to see individual rocks in a moving cement mixer.

It depends on what you're trying to accomplish. If all you want is GHz bandwidth over a short distance, such as for wireless HDMI TV, the antenna efficiency and gain are not terribly important. A conical monopole can easily have several octaves of bandwidth, if you don't mind the low gain.

True. However, low VSWR is the LEAST important specification for such a UWB antenna. From 3.1 to 10.6 GHz, it's no big deal.

However, if you're going to go outside the microwave regions, and into the "illegal" LF/MF/HF/VHF/UHF region, the antenna is going to be huge, ugly, inefficient, and probably expensive. Besides the FCC rules, there are other reasons to avoid these frequencies. The incumbent users of these frequencies do not appreciate spread spectrum gadgets raising their baseline noise level. Below about 30 Mhz, the atmospheric noise becomes sufficiently high that all you're going to hear is lightning strikes hundreds of miles away. I've done HF spread spectrum, but that was in a 30 KHz wide channel.

Nope. Loops are fairly narrow band and generally work with the magnetic field. A rod, I presume means a monopole, which is generally not a broadband structure. Multi-element is can be an LPDA (log periodic dipole array), which is broadband, but no fun to build at

10GHz. These are all suitable for VHF/UHF, but unlikely at microwave frequencies. Here's what UWB antennas look like:

Not synced but timed to avoid the RF bursts. The data gets through in between microwave oven bursts. Wi-Fi uses much the same idea.

--
Jeff Liebermann     jeffl@cruzio.com 
150 Felker St #D    http://www.LearnByDestroying.com 
Santa Cruz CA 95060 http://802.11junk.com 
Skype: JeffLiebermann     AE6KS    831-336-2558

Yeah, I may have done that with a spark gap transmitter I built when I was a kid. I didn't have much time to experiment because my father took it away when he discovered that it was illegal.

Well, I used my digits to send Morse code on a straight key. The same digits were used to operate the pencil I used to inscribe the message on paper. Since both ends used my digits, I guess it can be considered digital.

I think the quote was something like "Any sufficiently advanced communications technology is indistinguishable from noise". Only those with a broad base of knowledge and experience can understand that.

Auto manufacturers go through great lengths to pass FCC Part 15 incidental radiation requirements. Attaching an antenna to your vehicle ignition system is generally considered a bad idea. However, I used that to good effect many years ago. My commute home over Calif Hwy 9 involved many blind turns and homicidal drivers. Headlights could not be easily seen on the turns when there is a hill on the inside of the turn. So, I would tune my AM car radio to blank spot on the dial, and listen for the sound of ignition noise. Depending on the age of the vehicle, I could detect an oncoming vehicle at least

100ft ahead, and usually much farther.

The FCC is very picky about where one is allowed to experiment with new modulation and frequency generation schemes. At this time, it is impossible to legally build and test something that uses 0.1 MHz to

10,000 MHz. The best you can do today is UWB from 3.1 to 10.6 GHz:

--
Jeff Liebermann     jeffl@cruzio.com 
150 Felker St #D    http://www.LearnByDestroying.com 
Santa Cruz CA 95060 http://802.11junk.com 
Skype: JeffLiebermann     AE6KS    831-336-2558

When people talk about spread spectrum it is always in the context of transmitting data. But this is also used as a way to mitigate interference from unintentionally emitted noise. I have never seen data that actually indicates this works well. The most obvious use is dithering of a clock signal to "smear" out the noise spectrum. But in reality this doesn't reduce the level of the noise, it just moves about it so that most instruments won't measure the peak anymore. So the noise reduction is really just tricking the instruments and the peak value in the spectrum is still there. Whether or not it causes a problem with interference depends on the device being interfered with.

--

Rick

Before they can actually cause you any trouble they would have to measure your output to prove you are interfering. That sort of spread spectrum would likely have an amplitude measured by most test equipment that would not be appreciably above the noise, even at the output connector to the antenna. no?

--

Rick

Well, there's also the canola oil concoction called "Spectrum Spread".

I've demonstrated that I can move a CW signal through the bandpass of a Wi-Fi spread spectrum signal and there's little effect until the receiver front end overloads.

A more common use of spread spectrum is to reduce the difference between peak power and average power. Spreading out the signal reduces the peak power requirements for the RF power amplifier, and the dynamic range of the receiver chain. This has been in use since the "scrambler" was introduced in 1200 baud modems. AFAIK, this spreading of the signal to occupy as much of the allowed bandwidth as possible has been fundamental to every new communications technology since then.

The spread spectrum trick to pass Part 15 benefits from the FCC requiring measurement of only the peak signal power as seen on a spectrum analyzer. If those signals are spread over a wide range of frequencies, the peak power drops drastically. Without it, few of todays computah motherboards would pass.

--
Jeff Liebermann     jeffl@cruzio.com 
150 Felker St #D    http://www.LearnByDestroying.com 
Santa Cruz CA 95060 http://802.11junk.com 
Skype: JeffLiebermann     AE6KS    831-336-2558

The FCC enforcement burro is under DHS (Dept of Homeland Security) which probably doesn't really care, unless you are manufacturing a product or are acting on a complaint from a VIPC (very important political contributor). Convincing the FCC to enforce its own rules is a frustrating experience.

Most Wi-Fi and presumably UWB devices have a test connector on the PCB antenna connection to aid in production testing. This could be used to connect a spectrum analyzer or power meter to measure the output (-41.3 dBm/MHz). Or, just use a calibrated antenna. I presume that the devices have a 3.1 to 10.6 GHz stripline filter to eliminate out of band emissions:

--
Jeff Liebermann     jeffl@cruzio.com 
150 Felker St #D    http://www.LearnByDestroying.com 
Santa Cruz CA 95060 http://802.11junk.com 
Skype: JeffLiebermann     AE6KS    831-336-2558

Might have meant FHSS instead or as well.

?-)

Yes, that is what most people look at when doing spread spectrum by dithering a carrier. I am talking about the impact it has on *other* comms, essentially the EMI issue. People get gear through testing by dithering the clock but I don't know the mitigates the interference of other comms like TV, radio, etc. that is sensitive to narrow band interference.

Exactly. They pass, but do they actually interfere with other devices any less? The signal energy is actually spread across the spectrum only if you take a time average. The instantaneous energy at any given frequency is the same. I'm not sure this has a significantly smaller impact on the victim equipment or if it just "tricks" the test gear into averaging the energy over the spectrum. I believe the amount of dithering is actually pretty small, something of the order of 1%. This is large compared to the narrow bandwidth of the stationary peak, but well within the passband of most devices that could suffer interference. So the "averaging" likely has little impact in the real world.

--

Rick

You can have a frequency agile pre-filter with a hopper. So it is more jam resistant. But even without the frequency hopping pre-filter, the IF is narrower in a hopper, so there is less noise.

Ricochet gave a talk at the High Tech Experimenters Club back in the day. It would meet monthly at the old Ampex site. Google doesn't know of the clubs existance, so I guess finding a video of the session would be unlikely.

Saying "instantaneous energy" and "frequency" gets us into that tricky area. You can't really say that the energy is "at" any single frequency, at any given instant in time... "frequency" applies only over a period of time.

Consider the case of a really fast sweep, in which the "frequency" at the beginning of one sinewave cycle of the carrier is significantly different than the "frequencies" one cycle before or after. If you're constantly changing the clock rate this rapidly, it never really has time to "settle down" and deliver power into any one frequency, no?

I think it's going to depend on both the characteristics of the other equipment, and on the speed with which the clock signal is swept and the range over which it is swept.

In some cases, it probably makes a big difference (for one swept-spectrum interferer and one victim device). If the victim device has a narrow receive bandwidth, and the interferer is sweeping over a much larger bandwidth at a relatively high speed, then the victim device will see an occasional "chirp" sweep through its bandpass, but the interferer will spend most of its time "interfering elsewhere". Instead of the victim being wiped out by a strong "birdie" in its bandpass all of the time, it'll get a distributed "chirp" once in a while.

For wider-bandwidth victim devices, things aren't as rosy, since the interferer will be spending a lot of its time dumping energy

*somewhere* in the victim's passband.

Things are even less rosy when you talk about a whole bunch of interferers, each sweeping their clock rates around independently and incoherently. My impression is that you may avoid having a whole bunch of individual "birdy" interfering signals, but that you end up raising the overall noise floor of the whole band... all that noise power isn't just going away.

So, in that sense, you're right - the "spread spectrum" technique is a trick, which doesn't truly solve the underlying interference problem. I think it can make a modest number of interferers somewhat easier to deal with (raising broadband noise level, rather than emitting a bunch of strong narrow-bandwidth interference tones) but if you have enough systems using the same trick, the band noise becomes intolerable.