Advice on long distance wireless links?

Hiddink, Gerrit (Gerrit) ghiddink at agere.com
Mon Sep 9 17:31:14 EST 2002


Hi,

a problem you may be encountering is that the time the signal
takes to travel 45 kms exceeds the time that the peer waits for
an ACK. I'm currently too lazy to do these calculations for you,
I'm sorry. A MAC waits for SIFS time for an ACK. Some cards allow
this time to be changed.

You might consider testing a datarate lower than 11 Mbit/s.
Your SNR is 12, but might fluctuate quite a bit due to 
atmospheric conditions. With a lower datarate (and 
consequently a different modulation) you can improve
your chances of the signal being recognized from the noise.

You could try to set a lower fragmentation threshold to produce
smaller packets. Theoretically, smaller packets have a smaller
chance of "catching" a bit error. In practice, however, errors
come in bursts so that the fragmentation threshold may not help.

Regards, Gerrit

-----Original Message-----
From: Bruno Lopes F. Cabral [mailto:bruno at openline.com.br]
Sent: Monday, September 09, 2002 4:23 AM
To: wireless at lists.samba.org
Subject: Advice on long distance wireless links?


Hi there

I did a long distance point to point link here, 45km, 
30dBs 1.5meters antennaes on both sides (105m high on 
one side, 15m high on the other, no higher obstacles 
between the points)

the Link Quality shows 19/92, Signal level: -87 dBm 
Noise level:-99 dBm. auto bitrate states 1Mb, but 
fixing it at 11Mb also seems to "work" (i.e. pinging
the other side works -- most of the time)

the performance is poor in any case, about 64K/192K 
at best. lots of lost packets, very few or none DUP
packets

I dig a bit and found indicattions that long distance
links need increasing ACKs timeouts to prevent "multi-
path" retransmissions, but I could'nt find exactly what
I need to change or any further indications of how to
calculate the parameters involved, so I'm seeking for 
advice of you fellows

do I need to increase/decrease one or more of the 
following or other parameters? do you have any tables/
formulas indicating the ranges x the distances between 
the points?

- Fragmentation threeshold
- Listen interval (or beacon interval)
- RTS threeshold

would be better to use ad-hoc mode instead of managed 
mode in this setup?? I'm using linux on the stations,
tried orinoco (orinoco_cs/kernel 2.2) and prism2
(hostap_cs, kernel 2.4) cards (drivers)

any help/FAQ/pointers on this would be greatelly 
appreciated :)

Cheers
!3runo
from Brazil



More information about the wireless mailing list