Connect 2 buildings with wLan
Stafan Kahlert
I want to connect two appartments in 2 buildings with wireless lan.
The gap between the buildings is 50m max., with a tree inbetween.
If that tree wasn't there, we would have visual contact.
We planned to use 2 AccessPoints on either side as a bridge between
the two networks.
1.) Will this work?
2.) If yes, what a bandwith do we have to expect?
3.) If no, what equipment do we need?
thanks
S. Kahlert
David Taylor
Probably.
> 2.) If yes, what a bandwith do we have to expect?
A throughput of maybe up to 450KB/s, maybe more maybe less.
> 3.) If no, what equipment do we need?
Chainsaw.
The Team
Funny but true. LOS is needed in WLANS.
I would use two Buffalo Ap's (1)
(AirStation Standard Bridge Access Point )
two external cables (LRN400)
two pigtails
and two 24 dbi parabolic grids.
point them at each other and you should be hot to trot.
Tree would cause severe loss in streght but I think you would get
better than 450K. Matters how full tree is .Could rock in the winter
and suck swamp water in the summer.
Chainsaw best idea yet.
http://webplus.totalaccess.net
Kevin Webb
even a good one going through canopy. Now wait for a good wind storm. Then
see what you get. Then wait for a good rain. Deeee-ropped link. You can
just overpower the obstruction itself (probably anyway) with very high gain
directionals but the elements will getcha.
"Stafan Kahlert" <check...@t-online.de> wrote in message
news:alnl3s$vq6$01$1...@news.t-online.com...
David Taylor
50m and just one tree?!
Wow! I go through several trees and 200m with just a couple of tin cans
with no problems.
David.
Michael Erskine
> Funny but true. LOS is needed in WLANS.
> I would use two Buffalo Ap's (1)
> (AirStation Standard Bridge Access Point )
> two external cables (LRN400)
> two pigtails
> and two 24 dbi parabolic grids.
> point them at each other and you should be hot to trot.
Ye Gods!, Man, that setup is good for 6 to 8 *miles*.
It will work, just stick both AP's in windows on the same side of the
building, you'll get link.
Forget the grids, external pigtails, and the rest of the $500-$700 in
hardware, you definately will not need it.
-m-
Michael Erskine
> The problem with the tree is this, besides the obvious. Say you got a link,
> even a good one going through canopy. Now wait for a good wind storm. Then
> see what you get. Then wait for a good rain. Deeee-ropped link. You can
> just overpower the obstruction itself (probably anyway) with very high gain
> directionals but the elements will getcha.
>
Kevin;
He is asking about fifty yards between two buildings with one tree in
the way. The elements are not going to get him. He will have a
better than -80 dB receive signal just placing the units in windows
that face each other.
Hit the "So you think you understand ... " link. This is a FIFTY
meter link. He can probably put his transmit power at 10 mw on both
ends and do the link using the vendor supplied whips.
-m-
Kevin Webb
issue might not come into play but rain on the leaves will. 10 mW with
default dipoles in a window? Extremely unlikely.
"Michael Erskine" <osi...@deltaville.net> wrote in message
news:e59f93b2.02091...@posting.google.com...
David Taylor
> default dipoles in a window? Extremely unlikely.
I disagree. Maybe he'll need more than 10mW but a couple of tin cans
gets me 200m through several trees, reliably and in rain.
On it right now.
David.
Michael Erskine
While I respect your expertise:
I do agree with you that wet trees are worse than dry trees and that
wind blown wet trees are worse than dry still trees. It is only in
degree where we differ.
He's only going 150' feet. These units are rated in the 500-1000 foot
range given line of sight. Trees (vegetation) absorbs about 30 dB per
100 meters or say 8 to 20 dB per tree. If the receive sensitivity is
on the order of -76 DBm (matches the 802.11b spec) and the units have
an EIRP of 21 dBm (WAP-11 spec):
Path loss: -73.5 dB
Receive Ant: 2.1 dB
EIRP xmtr 21.0 dB
Tree loss: 20.0 dB
Receive Sens: 76.0 dB @ 11MBits
-73.5 + 76 + 2.1 + 21.0 - 20.0 = 5.6 dB fade margin at 11 MBits.
It looks like atleast a 5.5 MBit link to me. If he puts cookie sheets
behind both AP's he'll have a solid 15 + dB fade margin at 11 MBits,
unless the window screens are made of aluminium... or there is a metal
fence between them.
Therefore I feel you are in error on this one.
-m-
K Bloch
swat a fly with a sledgehammer. It may kill the fly but cause other
problems. A well done link provides for adequate signal at the
expected distance. Too much signal transmitted and received can also
be a problem. With the grid antennas you are likely to hear more
interference from others as well as having your signal propogate too
far. Do you really want people too miles away to be able to sniff all
the packets you send out over the air.
As far as throughput that depends on many things. From one PC to the
other it should be about 5 mbits per second max sustainable rate in
any one direction. If both users are trying to copy files from each
others systems the rate will probably drop to about 3 mbits each. If
you are talking about accessing the internet that will depend on the
connection speed you have to the internet. If the host network has a
dial up connection then you won't get any faster then that.
Line of sight is not always needed. the 2.4 ghz signal will penetrate
objects but it has its limits. 50 meters with a tree in the way may
not even need external antennas asuming the access points are placed
in windows that would otherwise have line of sight. I get nearly 400
feet from a cisco 1200 series access point indoors with supplied
antennas to a 350 client card thru trees and into a small concrete
block building. The connection is not rock solid but it works when
needed.
donot...@totalaccess.net (The Team) wrote in message news:<3d7faee8....@news.totalaccess.net>...
Kevin Webb
engineering but with hands-on experience. Fairly extensive as it is, I am
no expert. not by a long shot. So while the equation may work out, the
only way he'll know for sure is to try. Tree foliage is not an absolute so
numbers mean little in my opinion. He can try and I would suggest he do
that, but if it is a link that he must depend on then he should be prepared
for dissapointment, and hope for the best. Cheers.
"Michael Erskine" <osi...@deltaville.net> wrote in message
news:e59f93b2.02091...@posting.google.com...
Jeremy Parr
10mw or so. The tight beamwidth of the parabolics would REDUCE his chances
of interference and multipath IMHO.
Jermey
"K Bloch" <kbloc...@hotmail.com> wrote in message
news:26cfcb15.02091...@posting.google.com...
Michael Erskine
> Fair enough.
=) One of these days I am going to grow up.
[ snip ]
> So while the equation may work out, the only way he'll know for sure is to
> try.
Aye.
> Tree foliage is not an absolute so numbers mean little in my opinion.
Aye, tree foliage is a nightmare and very difficult to estimate with
even the slightest degree of accuracy. Even different species of
trees have different characterstics for different portions of the
spectrum. I don't think anyone has a good handle on it.
> He can try and I would suggest he do
> that, but if it is a link that he must depend on then he should be prepared
> for dissapointment, and hope for the best. Cheers.
That is always good advice. I really hate the disappointment part. I
have a 900 MHz link I am working with now. I tested the hardware to
1.6 miles at my house (thru substantial vegetation but little
construction) and then moved it to a setting in a small town. Link is
.5 miles thru a small town. No link, no sign of a signal. Walk the
path and "behold!" there is a large brick school house in the middle
of the path. Looks like ten feet on each end will fix it but to get
that I have to extend my power leads and if I do that I may as well go
up on the tower at the AP end. Please register my strong
disappointment. The good thing is that I will probably be able to
dump the one or both of the 13 dB panels in favor of a couple of
dipoles and thus remain legal for PtMP if I ever need it.
Cheers to you as well.
-m-
Don Widders
possible gain is that aiming the antennas can become difficult at extremely
high gain (well, that and 'ugly' antennas!)
One is likely to receive much less interference with high gain directional
antennas because they are less effective at picking up signals outside the
area of interest.
Security is usually improved using high gain directional antennas because
they do not transmit a signal in all directions. Reducing power to just
facilitate the desired link is an effective way to improve security.
Don W.
"K Bloch" <kbloc...@hotmail.com> wrote in message
news:26cfcb15.02091...@posting.google.com...