On Fri, 24 Oct 2008 01:29:44 -0700 (PDT),
mark.aren.nz@gmail.com
wrote:
>Up until a day or so ago I have been of the understanding that 802.11b
>used Direct Sequence Spread Spectrum (DSSS).
Nope. It can also be frequency hopping (FHSS). Products made by
Breezecom/Alvarion, Raylink, Proxim (Symphony), Symbol, and others
used FHSS. There are advantages to both technologies. FHSS never
went faster than 3Mbits/sec connection speed, receiver sensitivity was
limited, and the FCC rules restricted FHSS, resulting in lousy range
and speed. However, reliability and lack of susceptibility to
interference was a FHSS strength. Where DSSS and FHSS meet, FHSS just
slows down a little, while DSSS stops dead.
>I set up an Agilent Spectrum Analyser on the 2.400 - 2.450GHz band and
>was a little surprised to see that my 4 years old Access Point (Dlink
>DI-614+).appears to be using Frequency Hopping.
Nope. It's DSSS. If you see a classic sin(x)/x "hump" pattern on the
SA, it's DSSS. If it looks like a step functions, it's FHSS. You may
need ot adjust the sweep rate on the SA to see things clearly. Also
note that your DI-614+ is transmitting in bursts (probably just
beacons), which can look like FHSS is an exessively fast sweep. Move
some data through the wireless link and then watch the SA.
>From my understanding, FH uses a carrier which changes frequency 10s
>or 100s of times per second within it's allocated spectrum in a random
>manner.
Sorta. Bluegoof is 1600 hops per second. 802.11 FHSS is somewhat
faster, but I forgot the number (and am too lazy to RTFM for it).
>Spectral width of the carrier varies according to payload.
Nope. Occupied bandwidth is the same regardless of the traffic with
FHSS. The way it usually works is that data is sent with a 1 MHz wide
FM signal. There are 80 channels, each 1 MHz wide. FHSS is not
allowed to loiter on any one of these and must randomly hop through a
minimum of 75 channels before re-using any. What you'll see on the SA
should look like a 1 MHz wide signal moving over 83.5 MHz bandwidth.
>A DSSS signal in comparison should be noise like, with a reasonably
>rectangular spectrum.
Totally wrong. It should be a sin(x)/x envelope with a hump in the
middle. See:
<http://explorer.cyberstreet.com/wrt54g/WRT54g-spectraloutput.html>
Ugh. Those are awful photos but close enough to what you'll see with
uncontrolled input data. See photos at:
<http://sss-mag.com/ss.html>
for what DSSS and FHSS should look like. Also see the Yellowjacket
videos:
<http://www.bvsystems.com/Videos/videos.htm#yellowjacket-bang>
(after the sales pitch).
>From memory this is similar to ATSC TV in the US
>and CDMA telephony. Spectrum should not vary significantly with
>payload.
CDMA is spread spectrum and uses the same PN (pseudo noise) methods
used by DSSS. However, the modulation rates and spectral masks are
radically different. With DSSS, the air time (time the xmitter is on
the air) and modulation rate both vary with payload. I don't know
enough about ATSC to comment on it.
>We had some analog 2.4GHz phones a while ago and I do remember a
>continuous ticking noise in the background while the Wireless LAN was
>switched on (this matches the FH measurement).
That would be the 802.11 beacon broadcast used to identify the access
point.
>I would have expected a
>decrease in Signal to Noise ratio if the analog phone was close to a
>DSSS signal source.
Only if you were moving data on the 802.11 system. Also, you won't be
testing for SNR. It will be BER (bit error rate). It also depends on
the type of cordless phone. There are analog FM, DSSS, and FHSS
cordless phones with assorted compatibility issues. Some cordless
phone work cross band. It's impossible to predict the effects without
known the underlying technology and method of testing.
>Thoughts, comment and observations most welcome.
--
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
802.11b DSSS or FH ? 
Linear Mode
