HostAP digest, Vol 1 #61 - 14 msgs

Big Bloke bigbloke at sdf-eu.org
Tue Dec 10 07:12:04 EST 2002


On Mon, 9 Dec 2002 hostap-request at shmoo.com wrote:

Mike Wrote.

> I'm trying to understand how the tx/rx diversity works if anyone would
> be kind enough to enlighten me about their experiences I would really
> appreciate it.

Hi Mike,

as far as I know, (If I am wrong someone please correct me) TX diversity
is not used in WLAN.

The idea of RX diversity is to combat "Phasing" . If you have ever
listened to FM car radio under weak signal conditions you may have heard
"flutter" on the signal. Diversity is a method of countering
that effect.

When signals arrive at the receive antenna, they may do so from a
number of directions, either direct from the transmitter, or via
some object that has reflected the signal. If the signals arrive
together "In time" (i.e. in phase) then the two will add together
to produce a larger version of the original signal (additive mixing)
,if they arrive at different times, there is a chance that one of
the Signals will reduce the other (known as "subtractive mixing" )
and the resulting signal may be considerably weaker.

The two receive antennas are spaced a critical distance (a number
of wavelengths) apart in such a position, that if a signal is out
of phase with the one antenna, it should be in phase with the other.

Given that one wavelength is around 13cms then the spacing is quite
critical!

The Diversity circuit of the receiver will compare the phase difference
of the two received signals, delay the more advanced of the two so that
it is phased up with the other signal, then combine them together to
produce one composite stronger signal, which can be de-modulated more
effectively.

This technique has been used in the cellphone industry for around 20+
years.

IIRC, in a typical 2 antenna Access point setup, one antenna is both
TX and RX whereas the other is RX diversity only.

> Here is what I have tried so far:
> 1 -> Installed 2 Maxrad 10db omni's on standoff arms at about the top of
> a 100ft tower spaced 15ft apart horizontally, ran 2 separate Andrew's
> 1/2" Andrews Heliax cables to each antenna.

WOW! Serious setup ;-))

I thought my Homebrewed 28 Element omni into about 20 foot of LDF 5-50
was "Going some! " , But.... sheesh !!!  ;-)

The spacing of the two antennas is quite critical, as is the feeder
Length , which IIRC needs to be identical for the two antennas.
Also, considering the small power levels in use the
feeder length is rather long even for LDF/FSJ-450.

I calculate you would be losing 5.4 db Minimum climbing that 100Ft
Tower ! and 3db represents a 50% reduction in that signal level.
Is there anyway you could site the AP at the top of the tower and
just run Ethernet up to it ? There is commercial equipment available
to provide DC power for the AP over the Ethernet cable and the increase
in performance should be significant with only a few feet of heliax
between the AP and the Antennas.

<Snipped> !!

I'm not "quite" sure what you want to achieve with sectorising, I am
considering an OTSR solution (Omni Transmit, Sectored Receive) by
hacking into the WLAN card. to boost the "uplink" from the client cards.
as an alternative.

You might like to look at www.wimo.de who produce very good quality
"Power Splitters" for 2.4GHz that could be used for sectoring.

Any other questions, by all means mail me directly.

cheers

B B





More information about the HostAP mailing list