Extend wifi antenna range

[Paul]

In message <[email protected]>, Bill in Co
Paul wrote:
[]
Yes. Using an image search engine, I can find 24dBi parabolic Wifi
antennas, with something like a 7 degree beam width. So you can
web surf in your barn Many of them don't use a solid dish, and
instead use a "wire frame" reflector.
I guess that's what it's got to be for.
Or beaming it over to a distant room in a mansion, like the Hearst Castle.
Must be nice. Amazing.


Or, though of course the manufacturers don't actually push this in their
advertising!, clandestine activities, such as using a neighbour's (even
a distant one!) unsecured wifi, or eavesdropping with wireshark or
similar. I rather suspect such uses are responsible for more of the
sales of such things than just people with very big properties! (Though
some such - perhaps farms and the like, with widely-separated [and maybe
part-metal] outbuildings - I'm sure do exist.)

Here are a few http://ebay.eu/1jsgCww - though the little ones only
claim 8 dBi, so aren't too much better than the larger "rubber duck"
type; at these frequencies and for somethng reasonable in size, I think
yagi aerials http://ebay.eu/1gLcJkp still win (most of those are
claiming 25 dBi). Of course, the higher the gain, the narrower the beam,
so they're probably harder to point. Some of the bigger dishes on the
first link probably have comparable or even higher gains (and are even
harder to point).

IMO, a good yagi is the best.
At my borthers "ranch", he has a neighbor who uses a yagi to get a signal
because they live way off the grid.

One time as a trucker, I had a nice little experience with my littie
dongle.
I just toss the thing up on the dashboard for a good signal.
This particular night, I couldn't get a good enough signal to log on to
either of the two truckstops network.
But, there was a third signal and that one I logged on to.
That signal was coming from 25 mile up the road!
Albeit, it was a fluke due to the weather conditions.
Never got that luck again.

I live in Nova Scotia, Canada and our provincial government sent out a
contract a few years ago in order to insure everyone in province got
decent speed. Most rural areas receive this thru a network of wireless
transmission towers with the house using a yaga style attena pointed at
the tower. Works fine for me but I'm less than 1Km away from tower but
otherwise very rural. Only other option is dialup (slow) or sat
(limitations and expensive).

What kind of speed do you get with that ? Over 3Mbit/sec ?

*******

I found one more web page, with some comments on the Wifi time constants.

http://forums.wi-fiplanet.com/showthread.php?6488-what-is-the-time-out-value-in-802-11

slottime, ctstimeout, and acktimeout

Those sound like PHY level timers, rather than TCP/IP timers.
But a comment above that, mentions sending unacknowledged
packets, and that doesn't sound like a PHY function.

If someone has Wifi service, when they're more than
1.6KM away, there's got to be some reason for it that
I haven't been able to confirm yet.

If the tower was line of sight, and you have a laptop
with Wifi, perhaps you could drive further from the
tower, and see where the service "magically stops"
If at all.

Paul

 

[J. P. Gilliver (John)]

Shouldn't you be sending that story to Penthouse Forum?

He said it was 25 miles long ... (-:
--
J. P. Gilliver. UMRA: 1960/<1985 MB++G()AL-IS-Ch++(p)Ar@T+H+Sh0!:`)DNAf

"You realise, Fraser, that what happened between us can never repeat itself.
Unless, of course, the exact same circumstances were to repeat themselves." "By
exact same circumstances, sir, you mean: we would have to be aboard a train
loaded with unconscious Mounties, that had been taken over by terrorists, and
were heading for a nuclear catastrophe?" "Exactly." "Understood."


 

[mechanic]

In addition to power levels it is also important to
know how the 802.11 protocol acknowledge each received frame.
If the acknowledgement is not received, the frame is re-transmitted.

By default, the maximum distance between transmitter and receiver
is 1-mile (1.6 km).

As the signal would take only 5 microsecs or so to travel 1.6km,
that seems unlikely to be a factor.

 

[Dominique]

Paul <[email protected]> écrivait

What kind of speed do you get with that ? Over 3Mbit/sec ?

*******

I found one more web page, with some comments on the Wifi time constants.

http://forums.wi-fiplanet.com/showthread.php?6488-what-is-the-time-out- value-in-802-11

slottime, ctstimeout, and acktimeout

Those sound like PHY level timers, rather than TCP/IP timers.
But a comment above that, mentions sending unacknowledged
packets, and that doesn't sound like a PHY function.

If someone has Wifi service, when they're more than
1.6KM away, there's got to be some reason for it that
I haven't been able to confirm yet.

If the tower was line of sight, and you have a laptop
with Wifi, perhaps you could drive further from the
tower, and see where the service "magically stops"
If at all.

Paul

I was on a similar system before, I was getting between 3 and 5 Mbits/sec
but I was very close from the emitter. It was a Motorola Canopy wireless
system. But this system is not Wifi 802.11. You need a receiver (antenna)
at your end that you plug in the LAN port of a PC or the WAN port of a
router. So you cannot get the ISP signal only with a wifi laptop.

 

[Paul]

As the signal would take only 5 microsecs or so to travel 1.6km,
that seems unlikely to be a factor.

I can find articles that discuss setting the time constants.

http://www.air-stream.org/technical/ack-timeouts-and-effects-distance-links

I only brought this up, because it was mentioned in one of those
articles about the South America attempts to set distance records.
That article mentioned they had to tweak something in time
constants.

Paul

 

[Char Jackson]

A few years ago, Linksys opened up one of their routers to hacke...oops
experimenters. The 2.4GHz band is shared with an amateur radio
allocation, and hams began to modify the routers for all manner of
interesting uses. Hams are allowed to run much higher power than the puny
milliwatts the routers provide by default. And they did. High-gain
antennas and higher power allowed the routers to work over surprisingly
long distances. Some also modified the code to create automatic mesh
networks that were used for public service events like marathons.

I like how that story starts above, as if Linksys just decided to make the
source available out of the goodness of their hearts.

The way I remember it, they got caught using chunks of GPL code in their
router software and thus were forced, very much against their will, to make
the source available. And so were born third party firmwares that
immediately surpassed what Linksys had been able to do, such as dd-wrt (my
personal favorite), open-wrt, and others. In the past 5+ years, when I buy a
dd-wrt-compatible router, I immediately flash it before putting it into
service unless I find a compelling reason not to.

 

[Paul]

I can find articles that discuss setting the time constants.

http://www.air-stream.org/technical/ack-timeouts-and-effects-distance-links

I only brought this up, because it was mentioned in one of those
articles about the South America attempts to set distance records.
That article mentioned they had to tweak something in time
constants.

Paul

This link doesn't treat the timeout as an "absolute".

http://www.ab9il.net/wlan-projects/wifi1.html

"Bear in mind that beyond a distance of 3 km, throughput
may begin to decrease due to the ACK timeout setting of
the access point. ACK timing must be optimized for longer
distances - especially over links covering dozens of kilometers."

Any web sites from the major manufacturers, they draw
performance plots as if the thruput drops to zero
past a relatively short distance. I'd much rather
see an actual plot, maybe on a log scale or something.

Paul

 

[Jason]

I like how that story starts above, as if Linksys just decided to make the
source available out of the goodness of their hearts.

The way I remember it, they got caught using chunks of GPL code in their
router software and thus were forced, very much against their will, to make
the source available. And so were born third party firmwares that
immediately surpassed what Linksys had been able to do, such as dd-wrt (my
personal favorite), open-wrt, and others. In the past 5+ years, when I buy a
dd-wrt-compatible router, I immediately flash it before putting it into
service unless I find a compelling reason not to.

Interesting! I had no idea...

 

[pjp]

On Sat, 21 Dec 2013 08:45:10 +0000, J. P. Gilliver (John) wrote:

In message <[email protected]>, Bill in Co
Paul wrote:
[]
Yes. Using an image search engine, I can find 24dBi parabolic Wifi
antennas, with something like a 7 degree beam width. So you can
web surf in your barn Many of them don't use a solid dish, and
instead use a "wire frame" reflector.
I guess that's what it's got to be for.
Or beaming it over to a distant room in a mansion, like the Hearst Castle.
Must be nice. Amazing.


Or, though of course the manufacturers don't actually push this in their
advertising!, clandestine activities, such as using a neighbour's (even
a distant one!) unsecured wifi, or eavesdropping with wireshark or
similar. I rather suspect such uses are responsible for more of the
sales of such things than just people with very big properties! (Though
some such - perhaps farms and the like, with widely-separated [and maybe
part-metal] outbuildings - I'm sure do exist.)

Here are a few http://ebay.eu/1jsgCww - though the little ones only
claim 8 dBi, so aren't too much better than the larger "rubber duck"
type; at these frequencies and for somethng reasonable in size, I think
yagi aerials http://ebay.eu/1gLcJkp still win (most of those are
claiming 25 dBi). Of course, the higher the gain, the narrower the beam,
so they're probably harder to point. Some of the bigger dishes on the
first link probably have comparable or even higher gains (and are even
harder to point).
IMO, a good yagi is the best.
At my borthers "ranch", he has a neighbor who uses a yagi to get a signal
because they live way off the grid.

One time as a trucker, I had a nice little experience with my littie
dongle.
I just toss the thing up on the dashboard for a good signal.
This particular night, I couldn't get a good enough signal to log on to
either of the two truckstops network.
But, there was a third signal and that one I logged on to.
That signal was coming from 25 mile up the road!
Albeit, it was a fluke due to the weather conditions.
Never got that luck again.

I live in Nova Scotia, Canada and our provincial government sent out a

I often see 150+Kbs dl'ing a good torrent has lots of others.

Typical dl from someone like MS is approx 60-100Kbs. Dialup used to get
3.5Kbs and with sat it seemed like about 30Kbs (but that memory is
vague).

I'd expect to be able to play online games though I don't. Movies stream
fine for the most part. Weather has somewhat to do with it, better to
have clear shot at tower (I could shot it with a 22 here).

Service pack 3 for Xp took a couple hours for the 400megs or so.

It does seem to be affected by how many others are using the tower but
as I already said, I'm very rural so there can't be that many others on
it judging by how far apart the towers themselves are.

 

[J. P. Gilliver (John)]

As the signal would take only 5 microsecs or so to travel 1.6km,
that seems unlikely to be a factor.

That's ages in modern comm.s! The symbol rate at 1 Mb/s is 1 microsecond
- OK, somewhat longer for multiphase modulation, but 5 microseconds is
long enough to put a definite damper on throughput rate, by the time
you've allowed for symbol timing margins, handshake protocols, and all
the rest of the things you need. OK, you can design around it by use of
pipelining etc., which is what these people who experiment do, but the
basic protocol doesn't, as for short ranges that makes it unnecessarily
complex (and does reduce throughput).

 

[mechanic]

That's ages in modern comm.s! The symbol rate at 1 Mb/s is 1 microsecond

How big is the frame in the 802.11 protocol? Does each frame have to
be acknowledged before the next is transmitted? That might slow
things down, yes.