All day yesterday and through the night Team PAD
(http://www.wifiworldrecord.com) braved rain, lightning and winds over 30
mph to setup and test their equipment at their mountian top base outside of
Las Vegas, Nevada.
Today at 11am they successfully made a 125 mile link using 802.11b and ran
network applications with their remote team in the mountains West of St.
George, Utah.
More to follow, including photos, video and hardware details.
> All day yesterday and through the night Team PAD
> (http://www.wifiworldrecord.com) braved rain, lightning and winds over 30
> mph to setup and test their equipment at their mountian top base outside > of
> Las Vegas, Nevada.
Congrats to Team PAD. It's amazing that they established/held a link with those huge dishes in that wind. Besides the weather, it also looks like they had to contend with a couple peaks near Moapa River I.R. If they were much higher, it would've been a no-go.
Maybe they'll crush the Swedish 194mi amplified record before they return to OH and show them that Americans can do it without cheating. ;)
> All day yesterday and through the night Team PAD > (http://www.wifiworldrecord.com) braved rain, lightning and winds over 30 > mph to setup and test their equipment at their mountian top base outside of > Las Vegas, Nevada.
> Today at 11am they successfully made a 125 mile link using 802.11b and ran > network applications with their remote team in the mountains West of St. > George, Utah.
> More to follow, including photos, video and hardware details.
> Thank you
> Frank Keeney > Pasadena Networks, LLC > Antennas, Cables and Equipment: > http://www.wlanparts.com
"We also want to thank RF Linx for the use of amplifiers and pigtails"
So, Frank, on the World Record website the above shoutout to RF LINX appears
for letting them use AMPLIFIERS and pigtails. I thought this was supposed
to be an UNamplified demonstration? Am I thinking of the wrong one?
-----Original Message-----
From: SOCALWUG@googlegroups.com [mailto:SOCALWUG@googlegroups.com] On Behalf
Of Frank Keeney
Sent: Sunday, July 31, 2005 12:04 AM
To: sOCALWUG@googlegroups.com
Subject: [SOCALWUG] Defcon WiFi Shootout Record Set at 125 Miles for 802.11b
All day yesterday and through the night Team PAD
(http://www.wifiworldrecord.com) braved rain, lightning and winds over 30
mph to setup and test their equipment at their mountian top base outside of
Las Vegas, Nevada.
Today at 11am they successfully made a 125 mile link using 802.11b and ran
network applications with their remote team in the mountains West of St.
George, Utah.
More to follow, including photos, video and hardware details.
They had amplifiers for testing, but they were not needed for the 125 mile
link.
The only time they tried the amps was when the remote group was on a
mountain in Utah 145 miles away but could not quite reach the desired
coordinates 2 miles further and hundreds of feet higher, the road ended and
the terrain was too rough to continue driving. They were a little too low in
elevation, with a small mountain peak obstructing the two locations, they
tried unamplified and amplified, and the 802.11b link would not work.
If they made it to the desired location, I'm confident that they would have
successfully linked up at 145 miles.
> -----Original Message-----
> On Behalf Of Jim Sutton
> "We also want to thank RF Linx for the use of amplifiers and pigtails"
> So, Frank, on the World Record website the above shoutout to > RF LINX appears
> for letting them use AMPLIFIERS and pigtails. I thought this > was supposed
> to be an UNamplified demonstration? Am I thinking of the wrong one?
Last year, the team set up a link using amplifiers first, then disconnected
the amps to see if they would also be successful unamplified - they were.
Last year, they blew out the contest, winning amped, unamped, and style
categories. They probably used the same tactics this year, and wish to
credit the company.
So this year, again, the record is definitely for unamplified distance.
Wifi shootout coordinators call anything without an *external* amplifier,
"unamplified". So a Linksys adapter at 30mw, a Senao card at 200mw, or a
Zcom card at 300mw are all considered unamplified.
But if you use a box sitting between the card and the antenna boosting the
signal (whether it boosts from 30mw to 100mw or from 300 mw to 4 watts) it's
called "amplified" by the rules of the Defcon WiFi Shootout.
It remains to be seen what Guinness considers "unamplified" for sake of the
Guiness world record. (You may recall, the current Guiness amplified record
is 192 miles.)
-----Original Message-----
From: SOCALWUG@googlegroups.com [mailto:SOCALWUG@googlegroups.com] On Behalf
Of Jim Sutton
Sent: Sunday, July 31, 2005 11:38 AM
To: SOCALWUG@googlegroups.com
Subject: [SOCALWUG] They used AMPLIFIERS ?? RE: [SOCALWUG] Defcon WiFi
Shootout Record Set at 125 Miles for 802.11b
"We also want to thank RF Linx for the use of amplifiers and pigtails"
So, Frank, on the World Record website the above shoutout to RF LINX appears
for letting them use AMPLIFIERS and pigtails. I thought this was supposed
to be an UNamplified demonstration? Am I thinking of the wrong one?
Jim
-----Original Message-----
From: SOCALWUG@googlegroups.com [mailto:SOCALWUG@googlegroups.com] On Behalf
Of Frank Keeney
Sent: Sunday, July 31, 2005 12:04 AM
To: sOCALWUG@googlegroups.com
Subject: [SOCALWUG] Defcon WiFi Shootout Record Set at 125 Miles for 802.11b
All day yesterday and through the night Team PAD
(http://www.wifiworldrecord.com) braved rain, lightning and winds over 30
mph to setup and test their equipment at their mountian top base outside of
Las Vegas, Nevada.
Today at 11am they successfully made a 125 mile link using 802.11b and ran
network applications with their remote team in the mountains West of St.
George, Utah.
More to follow, including photos, video and hardware details.
If Team P.A.D.'s M.O. was the same as it was last year, maybe they used amps to ensure LOS and alignment, then established the link again after they removed the amps.
> "We also want to thank RF Linx for the use of amplifiers and pigtails"
> So, Frank, on the World Record website the above shoutout to RF LINX appears > for letting them use AMPLIFIERS and pigtails. I thought this was supposed > to be an UNamplified demonstration? Am I thinking of the wrong one?
> Jim
> -----Original Message----- > From: SOCALWUG@googlegroups.com [mailto:SOCALWUG@googlegroups.com] On Behalf > Of Frank Keeney > Sent: Sunday, July 31, 2005 12:04 AM > To: sOCALWUG@googlegroups.com > Subject: [SOCALWUG] Defcon WiFi Shootout Record Set at 125 Miles for 802.11b
> All day yesterday and through the night Team PAD > (http://www.wifiworldrecord.com) braved rain, lightning and winds over 30 > mph to setup and test their equipment at their mountian top base outside of > Las Vegas, Nevada.
> Today at 11am they successfully made a 125 mile link using 802.11b and ran > network applications with their remote team in the mountains West of St. > George, Utah.
> More to follow, including photos, video and hardware details.
> Thank you
> Frank Keeney > Pasadena Networks, LLC > Antennas, Cables and Equipment: > http://www.wlanparts.com
Frank: 145 miles away = Signal Peak? I thought they might try to go up there. Do you know if they still plan to shoot for the 192 mile amplified record before they return to OH?
> They had amplifiers for testing, but they were not needed for the 125 mile > link.
> The only time they tried the amps was when the remote group was on a > mountain in Utah 145 miles away but could not quite reach the desired > coordinates 2 miles further and hundreds of feet higher, the road ended and > the terrain was too rough to continue driving. They were a little too low in > elevation, with a small mountain peak obstructing the two locations, they > tried unamplified and amplified, and the 802.11b link would not work.
> If they made it to the desired location, I'm confident that they would have > successfully linked up at 145 miles.
> Thank you
> Frank Keeney > Pasadena Networks, LLC > Antennas, Cables and Equipment: > http://www.wlanparts.com
> > -----Original Message----- > > On Behalf Of Jim Sutton
> > "We also want to thank RF Linx for the use of amplifiers and pigtails"
> > So, Frank, on the World Record website the above shoutout to > > RF LINX appears > > for letting them use AMPLIFIERS and pigtails. I thought this > > was supposed > > to be an UNamplified demonstration? Am I thinking of the wrong one?
On Jul 31, 2005, at 2:25 AM, sherl...@comcast.net wrote:
> Congrats to Team PAD. It's amazing that they established/held a
> link with those huge dishes in that wind.
For a PTP link, the FCC says you can run 30dBm (1W) tx power into
6dBi of antenna gain, and then reduce the tx power 1dBm for every
3dBi of additional antenna gain. Here's a quick table of whats legal
in terms of EIRP:
The article says they were running "300mW" cards. Quoting:
"He tells me they used the VCom 325hp+ PCMCIA cards running at a
built-in power of 300 mw on each end of the link."
Lets give someone the benefit of the doubt, and say they were running
24.67dBm (293mW) into the antennas after the
pigtails and connectors. A quick look above will show that the
maximum allowed antenna gain (during transmit) is 22dBi.
Quoting again:
"The cards were connected to one 12 foot and one 10 foot diameter
satellite dish (see photo) on each side of the link."
Assuming that they got the engineering right, a 10' dish at 2400 MHz
would have 34.5 dBi of gain. At channel 11 it might have 35dBi.
Looking at the table above, you'll see that the most transmit power
they should be running with a 10' dish is 100mW (20dBm), so they're
way too high if they're using the 300mW cards. (* see below *)
So lets look at it another way. They claim 125 miles. The LOS path
loss for a 125 mile link using the 2.4 GHz is: 146 dB. So lets say
they use a 20dBm radio (to stay legal).
Lets also say they have perfect (zero loss) pigtails and cables, just
to keep it simple.
20dBm signal goes into 35dBi antenna, for 55dBm EIRP. The signal
then encounters 146 dB of path loss (at a minimum), so it
enters the other antenna at -91dBm, and then gets another 35dBi of
gain (due to the receive antenna), bringing the received signal
up to -56dBm, which is plenty high to decode an 11Mbps signal. (about 30dBm higher than required by a good radio).
Note that if they had a 25dBm signal (and wanted to stay EIRP legal)
the most antenna gain they could run would be 22 dBi. They have 47
dBm EIRP, which encounters the same 146 dB of path loss, for a signal
arriving at the remote receiver of -77 dBm, which is still "good
enough" to probably have a link at 11Mpbs.
However.... there is no way that 35dBi of antenna would pass FCC
testing on that card without at least 23dB of channel filter attached,
which they probably did not borrow or purchase. The antenna gain
raises the side lobes of the signal as well, and this will cause out
of band emissions, which is *bad*. The smaller antenna setup would
require a similar reduction in the size of the filtering required to
pass FCC.
They had other problems to overcome as well. 125 miles is 201km. The speed of light is 299,792.5 km/sec, so the signal takes 0.00067
sec to traverse the path (one way), or .00134 for a round trip.
When a packet is sent out from 802.11Station(A) it then waits for an
'ACKnowledgement frame' from 802.11Station(B). 802.11Station(A) will
only wait for a certain amount of time, this time is called the ACK
timeout or ACK window. If the ACK is NOT received within that timeout
period then the packet will be re-transmitted from 802.11Station(A)
resulting in reduced throughput.
Maximum theororetical ACK timeout for 802.11b is 744us or
approximately 111km /69 miles. (I could show the math, but you'd be
even more bored.) Their signal takes almost 2X the maximum ACK
timeout. Its just not fair that they called it "reliable", since
they're blowing several retransmissions on every attempted send, so
many, that I'm surprised the link works. One of the things I found
at Vivato was that
many, if not most cards would accept the first ACK even though it
arrives in the window when the second or third retransmit has been sent.
(And then I got to read the firmware source code for the Agere/ Intersil cards, and I understood 'why'.)
> For a PTP link, the FCC says you can run 30dBm (1W) tx power into
> 6dBi of antenna gain, and then reduce the tx power 1dBm for every
> 3dBi of additional antenna gain.
> Its just not fair that they called it "reliable", since
> they're blowing several retransmissions on every attempted send, so
> many, that I'm surprised the link works.
They deemed it "reliable" since it their applications, ssh to each other's
Linux laptop, VNC etc. were up and running uninterrupted for several hours
while they waited for the judges to arrive.
>> For a PTP link, the FCC says you can run 30dBm (1W) tx power into
>> 6dBi of antenna gain, and then reduce the tx power 1dBm for every
>> 3dBi of additional antenna gain.
> They ran their link under FCC Part 97 rules.
While Part 97 would give them the EIRP relief that you seem to seek
to justify, Part 97 also says "do not use more tx power than you need
to close the link" (not those exact words, but thats the jist of it),
and they're clearly running more than they need if they're 30dBm above
the sensitivity limit.
Further (and worse), Part 97 will *NOT* relieve them of transmitting
(spurious) signals in the restricted bands (well out of the Amateur bands
allowed under Part 97). This is still illegal, and any licensed HAM
would know better. Perhaps you will:
a) explain how they knocked the side lobes down so they didn't have
unwanted emissions outside the band. Your answer here very likely
has the word "filter" in it, but then I'll just ask you for the
mfg/model # of same, so I can verify its claimed performance.
b) explain how they transmitted their HAM call sign (perhaps they set
the ESSID to it, but this wouldn't fulfill the requirements of Part 97)
c) send along the call sign(s) used during this experiment, so the
FCC can pull their tickets.
> They deemed it "reliable" since it their applications, ssh to each
> other's
> Linux laptop, VNC etc. were up and running uninterrupted for
> several hours
> while they waited for the judges to arrive.
They deemed it reliable because they wanted to. Did they put any
load on the network? Many people will misinterpret the results,
and clamor for 125 mile links with no other requirements than "line
of sight" and "big ears". For the reasons I've shown, and others,
while you might be able to "show" this in a "demo", its both illegal,
and it doesn't (nee: can't) work correctly.
Netgate also gets requests to support these types of events with the
ever-present justification, "free advertising". We always turn them
down, because I don't want my name sullied with activities in the
802.11 space that are of questionable legality.
Jim
p.s. Vivato, when it was Mabuhay, attached an ordinary Lucent PCMCIA
card to a 12' dish (with the appropriate cavity filter in-place) and
got bits to/from someone's laptop (also with a Lucent card) who was
sitting on the chairlift some 12 miles away. I mention this because,
other than aiming the dishes, the Shootout isn't really "new", and I
happen to know that cavity filters exist that will do the job (as
well as what they cost.)
-----Original Message-----
From: SOCALWUG@googlegroups.com [mailto:SOCALWUG@googlegroups.com] On Behalf
Of Jim Thompson
Sent: Monday, August 01, 2005 12:15 AM
To: SOCALWUG@googlegroups.com
Subject: [SOCALWUG] Re: Defcon WiFi Shootout Record Set at 125 Miles for
802.11b
On Jul 31, 2005, at 2:25 AM, sherl...@comcast.net wrote:
> Congrats to Team PAD. It's amazing that they established/held a
> link with those huge dishes in that wind.
For a PTP link, the FCC says you can run 30dBm (1W) tx power into
6dBi of antenna gain, and then reduce the tx power 1dBm for every
3dBi of additional antenna gain. Here's a quick table of whats legal
in terms of EIRP:
The article says they were running "300mW" cards. Quoting:
"He tells me they used the VCom 325hp+ PCMCIA cards running at a
built-in power of 300 mw on each end of the link."
Lets give someone the benefit of the doubt, and say they were running
24.67dBm (293mW) into the antennas after the
pigtails and connectors. A quick look above will show that the
maximum allowed antenna gain (during transmit) is 22dBi.
Quoting again:
"The cards were connected to one 12 foot and one 10 foot diameter
satellite dish (see photo) on each side of the link."
Assuming that they got the engineering right, a 10' dish at 2400 MHz
would have 34.5 dBi of gain. At channel 11 it might have 35dBi.
Looking at the table above, you'll see that the most transmit power
they should be running with a 10' dish is 100mW (20dBm), so they're
way too high if they're using the 300mW cards. (* see below *)
So lets look at it another way. They claim 125 miles. The LOS path
loss for a 125 mile link using the 2.4 GHz is: 146 dB. So lets say
they use a 20dBm radio (to stay legal).
Lets also say they have perfect (zero loss) pigtails and cables, just
to keep it simple.
20dBm signal goes into 35dBi antenna, for 55dBm EIRP. The signal
then encounters 146 dB of path loss (at a minimum), so it
enters the other antenna at -91dBm, and then gets another 35dBi of
gain (due to the receive antenna), bringing the received signal
up to -56dBm, which is plenty high to decode an 11Mbps signal. (about 30dBm higher than required by a good radio).
Note that if they had a 25dBm signal (and wanted to stay EIRP legal)
the most antenna gain they could run would be 22 dBi. They have 47
dBm EIRP, which encounters the same 146 dB of path loss, for a signal
arriving at the remote receiver of -77 dBm, which is still "good
enough" to probably have a link at 11Mpbs.
However.... there is no way that 35dBi of antenna would pass FCC
testing on that card without at least 23dB of channel filter attached,
which they probably did not borrow or purchase. The antenna gain
raises the side lobes of the signal as well, and this will cause out
of band emissions, which is *bad*. The smaller antenna setup would
require a similar reduction in the size of the filtering required to
pass FCC.
They had other problems to overcome as well. 125 miles is 201km. The speed of light is 299,792.5 km/sec, so the signal takes 0.00067
sec to traverse the path (one way), or .00134 for a round trip.
When a packet is sent out from 802.11Station(A) it then waits for an
'ACKnowledgement frame' from 802.11Station(B). 802.11Station(A) will
only wait for a certain amount of time, this time is called the ACK
timeout or ACK window. If the ACK is NOT received within that timeout
period then the packet will be re-transmitted from 802.11Station(A)
resulting in reduced throughput.
Maximum theororetical ACK timeout for 802.11b is 744us or
approximately 111km /69 miles. (I could show the math, but you'd be
even more bored.) Their signal takes almost 2X the maximum ACK
timeout. Its just not fair that they called it "reliable", since
they're blowing several retransmissions on every attempted send, so
many, that I'm surprised the link works. One of the things I found
at Vivato was that
many, if not most cards would accept the first ACK even though it
arrives in the window when the second or third retransmit has been sent.
(And then I got to read the firmware source code for the Agere/ Intersil cards, and I understood 'why'.)
jim
-- No virus found in this incoming message.
Checked by AVG Anti-Virus.
Version: 7.0.338 / Virus Database: 267.9.8/61 - Release Date: 8/1/2005
-- No virus found in this outgoing message.
Checked by AVG Anti-Virus.
Version: 7.0.338 / Virus Database: 267.9.8/61 - Release Date: 8/1/2005
Jim, Ham radio regulations (Part 97) differ quite a bit from unlicensed regulations (Part 15)...
Jim Thompson wrote:
<snip>
> a) explain how they knocked the side lobes down so they didn't have > unwanted emissions outside the band. Your answer here very likely > has the word "filter" in it, but then I'll just ask you for the > mfg/model # of same, so I can verify its claimed performance.
Under part 97.305, amateur radio stations are allowed to use spread spectrum in the 13cm band 2390-2450 MHz. The signal has to be 802.11b, not 802.11g encoded (OFDM is not Spread Spectrum). Amateurs are allowed to use up to 1 watt without automatic power control and up to 100 watts with APC. There is NO limit on antenna gain. The ARRL bandplan is a suggestion, not an FCC requirement.
Amateurs can operate on channels 1-6 using part 15 equipment. Typically hams use channel 5 or 2432 MHz as a center frequency to avoid interfering with amateur satellite transmissions near channel 1 or 2400-2410 Mhz. Furthermore, by picking channel 5, you have 18 Mhz above and 42 Mhz below center frequency for the +/- 11 Mhz signal to fall into. Part 97.307 describes the spurious emissions requirements.
>From what I can tell, their setup could have met these requirements. > b) explain how they transmitted their HAM call sign (perhaps they set > the ESSID to it, but this wouldn't fulfill the requirements of Part 97)
Amateur stations using 802.11b must identify every 10 minutes according to Part 97.119 b(3). ASCII is a recognized code. Setting the SSID to callsign1-callsign2 in the peer-to-peer mode is enough for ID. However, it is also acceptable to send identification in the text of the message, so that anyone with a normal 802.11b card can read the identities of the transmitting stations. I believe that if they followed either of these methods, their setup would have met the regs.
> c) send along the call sign(s) used during this experiment, so the > FCC can pull their tickets.
Please tell me where you think they broke any regulations?
Their setup seems to use the minimum required power required to maintain the link. The high gain dishes minimize interference with others using the same band.
But out of curiosity, I too would like to know their callsigns and would be interested in knowing if they met the part 97 logging rule of their contact.
<snip>
> They deemed it reliable because they wanted to. Did they put any > load on the network? Many people will misinterpret the results, > and clamor for 125 mile links with no other requirements than "line > of sight" and "big ears". For the reasons I've shown, and others, > while you might be able to "show" this in a "demo", its both illegal, > and it doesn't (nee: can't) work correctly.
Give these guys a break. These guys are hackers. Who says it has to be stock software. They should have been able to hack the retry timer value in the firmware.
For a valid contact on ham radio, they have to exchange callsigns and a signal report. There is no requirement to have a network load to make a contact.
this doesn't mean they're FCC compliant (even if it did happen). Y
Which it probably didn't.
You still have 35 dBi of antenna gain raising the side lobes.
http://www.wlanparts.com/product/XI-325HPPLUS says "up to
24.7dBm" (292mW), and since these are Intersil/connexant Prism2- based, you can't "tune the power" like you could with an atheros- based card.
On Aug 1, 2005, at 9:20 PM, Humphrey Cheung wrote:
> -----Original Message-----
> From: SOCALWUG@googlegroups.com [mailto:SOCALWUG@googlegroups.com]
> On Behalf
> Of Jim Thompson
> Sent: Monday, August 01, 2005 12:15 AM
> To: SOCALWUG@googlegroups.com
> Subject: [SOCALWUG] Re: Defcon WiFi Shootout Record Set at 125
> Miles for
> 802.11b
> On Jul 31, 2005, at 2:25 AM, sherl...@comcast.net wrote:
>> Congrats to Team PAD. It's amazing that they established/held a
>> link with those huge dishes in that wind.
> For a PTP link, the FCC says you can run 30dBm (1W) tx power into
> 6dBi of antenna gain, and then reduce the tx power 1dBm for every
> 3dBi of additional antenna gain. Here's a quick table of whats legal
> in terms of EIRP:
> The article says they were running "300mW" cards. Quoting:
> "He tells me they used the VCom 325hp+ PCMCIA cards running at a
> built-in power of 300 mw on each end of the link."
> Lets give someone the benefit of the doubt, and say they were running
> 24.67dBm (293mW) into the antennas after the
> pigtails and connectors. A quick look above will show that the
> maximum allowed antenna gain (during transmit) is 22dBi.
> Quoting again:
> "The cards were connected to one 12 foot and one 10 foot diameter
> satellite dish (see photo) on each side of the link."
> Assuming that they got the engineering right, a 10' dish at 2400 MHz
> would have 34.5 dBi of gain. At channel 11 it might have 35dBi.
> Looking at the table above, you'll see that the most transmit power
> they should be running with a 10' dish is 100mW (20dBm), so they're
> way too high if they're using the 300mW cards. (* see below *)
> So lets look at it another way. They claim 125 miles. The LOS path
> loss for a 125 mile link using the 2.4 GHz is: 146 dB. So lets say
> they use a 20dBm radio (to stay legal).
> Lets also say they have perfect (zero loss) pigtails and cables, just
> to keep it simple.
> 20dBm signal goes into 35dBi antenna, for 55dBm EIRP. The signal
> then encounters 146 dB of path loss (at a minimum), so it
> enters the other antenna at -91dBm, and then gets another 35dBi of
> gain (due to the receive antenna), bringing the received signal
> up to -56dBm, which is plenty high to decode an 11Mbps signal.
> (about 30dBm higher than required by a good radio).
> Note that if they had a 25dBm signal (and wanted to stay EIRP legal)
> the most antenna gain they could run would be 22 dBi. They have 47
> dBm EIRP, which encounters the same 146 dB of path loss, for a signal
> arriving at the remote receiver of -77 dBm, which is still "good
> enough" to probably have a link at 11Mpbs.
> However.... there is no way that 35dBi of antenna would pass FCC
> testing on that card without at least 23dB of channel filter attached,
> which they probably did not borrow or purchase. The antenna gain
> raises the side lobes of the signal as well, and this will cause out
> of band emissions, which is *bad*. The smaller antenna setup would
> require a similar reduction in the size of the filtering required to
> pass FCC.
> They had other problems to overcome as well. 125 miles is 201km.
> The speed of light is 299,792.5 km/sec, so the signal takes 0.00067
> sec to traverse the path (one way), or .00134 for a round trip.
> When a packet is sent out from 802.11Station(A) it then waits for an
> 'ACKnowledgement frame' from 802.11Station(B). 802.11Station(A) will
> only wait for a certain amount of time, this time is called the ACK
> timeout or ACK window. If the ACK is NOT received within that timeout
> period then the packet will be re-transmitted from 802.11Station(A)
> resulting in reduced throughput.
> Maximum theororetical ACK timeout for 802.11b is 744us or
> approximately 111km /69 miles. (I could show the math, but you'd be
> even more bored.) Their signal takes almost 2X the maximum ACK
> timeout. Its just not fair that they called it "reliable", since
> they're blowing several retransmissions on every attempted send, so
> many, that I'm surprised the link works. One of the things I found
> at Vivato was that
> many, if not most cards would accept the first ACK even though it
> arrives in the window when the second or third retransmit has been
> sent.
> (And then I got to read the firmware source code for the Agere/
> Intersil cards, and I understood 'why'.)
> jim
> -- > No virus found in this incoming message.
> Checked by AVG Anti-Virus.
> Version: 7.0.338 / Virus Database: 267.9.8/61 - Release Date: 8/1/2005
> -- > No virus found in this outgoing message.
> Checked by AVG Anti-Virus.
> Version: 7.0.338 / Virus Database: 267.9.8/61 - Release Date: 8/1/2005
I'm more inclined to trust these kids, who have a proven track record. In
addition, Frank was up there with them.
You keep asking for their call signs... if you did 15 seconds of searching,
it's on the web. The call signs are on their website, and on pictures of
the dish,.
With hacked HostAP drivers, you can tune the power of Prism 2 cards.
Here is one example.
------------------------------------
To control automatic controlling of transmission power ‘/sbin/iwpriv wlan0
prims2_param 5 X’ , where X = 0 for disabling and 1 for enabling To read the current value of the transmission power ‘/sbin/iwpriv wlan0
readmif 62’ To change the current value of the transmission power ‘/sbin/iwpriv wlan0
prism2_param 6 Y’ , where 0 <= Y <= 255 (value for min power level = 127 , value for max power level = 128)
------------------------------------
-----Original Message-----
From: SOCALWUG@googlegroups.com [mailto:SOCALWUG@googlegroups.com] On Behalf
Of Jim Thompson
Sent: Tuesday, August 02, 2005 1:35 AM
To: SOCALWUG@googlegroups.com
Subject: [SOCALWUG] Re: Defcon WiFi Shootout Record Set at 125 Miles for
802.11b
this doesn't mean they're FCC compliant (even if it did happen). Y
Which it probably didn't.
You still have 35 dBi of antenna gain raising the side lobes.
http://www.wlanparts.com/product/XI-325HPPLUS says "up to
24.7dBm" (292mW), and since these are Intersil/connexant Prism2- based, you can't "tune the power" like you could with an atheros- based card.
On Aug 1, 2005, at 9:20 PM, Humphrey Cheung wrote:
> I interviewed the team after the awards ceremony.
> -----Original Message-----
> From: SOCALWUG@googlegroups.com [mailto:SOCALWUG@googlegroups.com]
> On Behalf
> Of Jim Thompson
> Sent: Monday, August 01, 2005 12:15 AM
> To: SOCALWUG@googlegroups.com
> Subject: [SOCALWUG] Re: Defcon WiFi Shootout Record Set at 125
> Miles for
> 802.11b
> On Jul 31, 2005, at 2:25 AM, sherl...@comcast.net wrote:
>> Congrats to Team PAD. It's amazing that they established/held a
>> link with those huge dishes in that wind.
> For a PTP link, the FCC says you can run 30dBm (1W) tx power into
> 6dBi of antenna gain, and then reduce the tx power 1dBm for every
> 3dBi of additional antenna gain. Here's a quick table of whats legal
> in terms of EIRP:
> The article says they were running "300mW" cards. Quoting:
> "He tells me they used the VCom 325hp+ PCMCIA cards running at a
> built-in power of 300 mw on each end of the link."
> Lets give someone the benefit of the doubt, and say they were running
> 24.67dBm (293mW) into the antennas after the
> pigtails and connectors. A quick look above will show that the
> maximum allowed antenna gain (during transmit) is 22dBi.
> Quoting again:
> "The cards were connected to one 12 foot and one 10 foot diameter
> satellite dish (see photo) on each side of the link."
> Assuming that they got the engineering right, a 10' dish at 2400 MHz
> would have 34.5 dBi of gain. At channel 11 it might have 35dBi.
> Looking at the table above, you'll see that the most transmit power
> they should be running with a 10' dish is 100mW (20dBm), so they're
> way too high if they're using the 300mW cards. (* see below *)
> So lets look at it another way. They claim 125 miles. The LOS path
> loss for a 125 mile link using the 2.4 GHz is: 146 dB. So lets say
> they use a 20dBm radio (to stay legal).
> Lets also say they have perfect (zero loss) pigtails and cables, just
> to keep it simple.
> 20dBm signal goes into 35dBi antenna, for 55dBm EIRP. The signal
> then encounters 146 dB of path loss (at a minimum), so it
> enters the other antenna at -91dBm, and then gets another 35dBi of
> gain (due to the receive antenna), bringing the received signal
> up to -56dBm, which is plenty high to decode an 11Mbps signal.
> (about 30dBm higher than required by a good radio).
> Note that if they had a 25dBm signal (and wanted to stay EIRP legal)
> the most antenna gain they could run would be 22 dBi. They have 47
> dBm EIRP, which encounters the same 146 dB of path loss, for a signal
> arriving at the remote receiver of -77 dBm, which is still "good
> enough" to probably have a link at 11Mpbs.
> However.... there is no way that 35dBi of antenna would pass FCC
> testing on that card without at least 23dB of channel filter attached,
> which they probably did not borrow or purchase. The antenna gain
> raises the side lobes of the signal as well, and this will cause out
> of band emissions, which is *bad*. The smaller antenna setup would
> require a similar reduction in the size of the filtering required to
> pass FCC.
> They had other problems to overcome as well. 125 miles is 201km.
> The speed of light is 299,792.5 km/sec, so the signal takes 0.00067
> sec to traverse the path (one way), or .00134 for a round trip.
> When a packet is sent out from 802.11Station(A) it then waits for an
> 'ACKnowledgement frame' from 802.11Station(B). 802.11Station(A) will
> only wait for a certain amount of time, this time is called the ACK
> timeout or ACK window. If the ACK is NOT received within that timeout
> period then the packet will be re-transmitted from 802.11Station(A)
> resulting in reduced throughput.
> Maximum theororetical ACK timeout for 802.11b is 744us or
> approximately 111km /69 miles. (I could show the math, but you'd be
> even more bored.) Their signal takes almost 2X the maximum ACK
> timeout. Its just not fair that they called it "reliable", since
> they're blowing several retransmissions on every attempted send, so
> many, that I'm surprised the link works. One of the things I found
> at Vivato was that
> many, if not most cards would accept the first ACK even though it
> arrives in the window when the second or third retransmit has been
> sent.
> (And then I got to read the firmware source code for the Agere/
> Intersil cards, and I understood 'why'.)
> jim
> -- > No virus found in this incoming message.
> Checked by AVG Anti-Virus.
> Version: 7.0.338 / Virus Database: 267.9.8/61 - Release Date: 8/1/2005
> -- > No virus found in this outgoing message.
> Checked by AVG Anti-Virus.
> Version: 7.0.338 / Virus Database: 267.9.8/61 - Release Date: 8/1/2005
-- No virus found in this incoming message.
Checked by AVG Anti-Virus.
Version: 7.0.338 / Virus Database: 267.9.8/61 - Release Date: 8/1/2005
-- No virus found in this outgoing message.
Checked by AVG Anti-Virus.
Version: 7.0.338 / Virus Database: 267.9.8/61 - Release Date: 8/1/2005
Jim Thompson wrote: > this doesn't mean they're FCC compliant (even if it did happen). Y
What is not FCC compliant?
> Which it probably didn't.
It's heresay on your part as well as theirs. All I am showing you is that what they are telling you is legally possible under FCC part 97 rules.
> You still have 35 dBi of antenna gain raising the side lobes.
Under part 97 rules, spurious emssions are measured relative to the input power to the antenna of the fundamental emission. The antenna gain applies to both the main carrier and the spurs.
97.307(e) "For a transmitter having a mean power of 25 W or less, the mean power of any spurious emission supplied to the antenna transmission line must not exceed 25 uW and must be at least 40 dB below the mean power of the fundamental emission, but need not be reduced below the power of 10 uW."
> http://www.wlanparts.com/product/XI-325HPPLUS says "up to > 24.7dBm" (292mW), and since these are Intersil/connexant Prism2- > based, you can't "tune the power" like you could with an atheros- > based card.
The Prism 2.5 Chipset can be power controlled under Linux. I have done it on a Senao 2511, but not this Z-Com card.
> On Aug 1, 2005, at 9:20 PM, Humphrey Cheung wrote:
> > I interviewed the team after the awards ceremony.
> Jim,
> Ham radio regulations (Part 97) differ quite a bit from unlicensed
> regulations (Part 15)...
> Jim Thompson wrote:
> <snip>
>> a) explain how they knocked the side lobes down so they didn't have
>> unwanted emissions outside the band. Your answer here very likely
>> has the word "filter" in it, but then I'll just ask you for the
>> mfg/model # of same, so I can verify its claimed performance.
> Under part 97.305, amateur radio stations are allowed to use spread
> spectrum in the 13cm band 2390-2450 MHz.
Understood (Kd5FGA, btw).
> The signal has to be 802.11b, not 802.11g encoded (OFDM is not
> Spread Spectrum).
Yeah, but to be pedantic, it could be 802.11 DSSS as well. (802.11b
requires CCK modulation.)
> Amateurs are allowed to use up to 1 watt without automatic power
> control and up to
> 100 watts with APC.
Yep.
> There is NO limit on antenna gain.
As long as you don't emit signals outside the band above the (strict
in this case) established limits.
> The ARRL bandplan is a suggestion, not an FCC requirement.
this is a true statement, but seems to walk over the FCC requirements
as well. Amateur operators can't transmit anywhere
they wish.
> Amateurs can operate on channels 1-6 using part 15 equipment.
Yep
> Typically hams use channel 5 or 2432 MHz as a center frequency to
> avoid
> interfering with amateur satellite transmissions near channel 1 or
> 2400-2410 Mhz. Furthermore, by picking channel 5, you have 18 Mhz
> above and 42 Mhz below center frequency for the +/- 11 Mhz signal to
> fall into. Part 97.307 describes the spurious emissions requirements.
Right, and have you looked at the sin(x)/x characteristics of the
signal? Have you looked at the band edge
performance plots available in the FCC filings for this card?
Remember that the band edge limits are strict limits (even for
HAMs). This is where the huge antenna gain will get you in trouble.
>> From what I can tell, their setup could have met these requirements.
Unlikely without some fairly big channel (or band) filters.
>> b) explain how they transmitted their HAM call sign (perhaps they set
>> the ESSID to it, but this wouldn't fulfill the requirements of
>> Part 97)
> Amateur stations using 802.11b must identify every 10 minutes
> according
> to Part 97.119 b(3). ASCII is a recognized code. Setting the SSID to
> callsign1-callsign2 in the peer-to-peer mode is enough for ID.
> However, it is also acceptable to send identification in the text of
> the message, so that anyone with a normal 802.11b card can read the
> identities of the transmitting stations. I believe that if they
> followed either of these methods, their setup would have met the regs.
Yeah, OK. Are *they* going to explain?
>> c) send along the call sign(s) used during this experiment, so the
>> FCC can pull their tickets.
> Please tell me where you think they broke any regulations?
any number of potentials exist. To be perfectly blunt, I think the
"we operated under part 97" was a dodge.
> Their setup seems to use the minimum required power required to
> maintain the link.
I think I showed that the EIRP was at least 25dB over the 11Mbps
sensitivity level. Thats a *BUNCH* of fade margin.
> The high gain dishes minimize interference with others using the
> same band.
Its unlikely there were any other users in the line of sight. I
accept that.
> But out of curiosity, I too would like to know their callsigns and
> would be interested in knowing if they met the part 97 logging rule of
> their contact.
> <snip>
>> They deemed it reliable because they wanted to. Did they put any
>> load on the network? Many people will misinterpret the results,
>> and clamor for 125 mile links with no other requirements than "line
>> of sight" and "big ears". For the reasons I've shown, and others,
>> while you might be able to "show" this in a "demo", its both illegal,
>> and it doesn't (nee: can't) work correctly.
> Give these guys a break. These guys are hackers. Who says it has
> to be stock software. They should have been able to hack the retry
> timer value in the firmware.
Nope. About all you can do is change the number of retries on that f/w.
> For a valid contact on ham radio, they have to exchange callsigns
> and a
> signal report. There is no requirement to have a network load to make
> a contact.
Violation of FCC part 97 rules? ---> "Both ends of the link were using
SSH"
97.309(b) (b) Where authorized by Secs. 97.305(c) and 97.307(f) of this
part, a station may transmit a RTTY or data emission using an
unspecified digital code, except to a station in a country with which
the United States does not have an agreement permitting the code to be
used. RTTY and data emissions using unspecified digital codes must not
be transmitted for the purpose of obscuring the meaning of any
communication. When deemed necessary by an EIC to assure compliance
with the FCC Rules, a station must:
(1) Cease the transmission using the unspecified digital code;
(2) Restrict transmissions of any digital code to the extent
instructed;
(3) Maintain a record, convertible to the original information, of
all digital communications transmitted.
Why did they not open a Telnet session that is open for every observer
to see? Why did they need an SSH tunnel on this link?
"The connection had an astounding signal strength of -37 dBm, bottoming
out at around -50 dBm throughout the 3 hours. The noise levels were
around -84 dBm"
With that kind of link margin, they should have been able to do this
with much more power. (6 dB = double the distance)
What on earth do they need an amp for??
Violation of FCC part 97 rules? ---> "When the amp was used the signal
strength bumped up to -18 dBm at 124.9
miles!"
97.313 Transmitter power standards.
(a) An amateur station must use the minimum transmitter power
necessary
to carry out the desired communications.
also interesting...
"the base station end had FAA 2.4 ghz communications so we chose to
keep our ERP low on that end"
Why did they need an unbalanced link??
Jim,
To be perfectly fair about it, I think that DefCon should amend their
rules for the shootout to limit it under FCC part 15 regs and any stock
Wi-Fi card. Without this restriction, the DefCon shootout is
meaningless. I don't have any problem with enhancing the drivers do
deal with the distance, but it will just become a my dish is bigger
than your dish contest. ...I wonder if doppler is too much to make an
EME contact using Wi-Fi?
Sure you can, the linux command is iwconfig wlan0 txpower <whatever between -43 (off) and +24> or "auto", this works fine wiuth prism 2.5 or 3 cards including zcomax
> Violation of FCC part 97 rules? ---> "Both ends of the link were
> using SSH"
Yeah, thats probably another violation.
> "The connection had an astounding signal strength of -37 dBm,
> bottoming
> out at around -50 dBm throughout the 3 hours. The noise levels were
> around -84 dBm"
> With that kind of link margin, they should have been able to do this
> with much more power. (6 dB = double the distance)
This is only true when the path loss coefficient is 2.0 (perfect
LOS). Jim's quick rule of thumb:
divide the delta link margin (in dB) by the path loss coefficient,
and the result is the difference in distanced, expressed in dB.
So, 6 dB / 2.0 = 3 dB, or 2X the distance.
Anyway, you're right, but only for LOS communications.
Further, something is ... spooky.
The 10' dish has a gain of around 35 dBi.
A 1 m^2 dish will generate about 26-27 dBi. 3 dBi more requires
twice the "aperture", or 2 m^2 for 29dBi 32 dBi would require 4 m^2,
and 35 dBi would require 8 m^2. There are 10.76 ft^2 in a square
meter.
(As an aside, consider just how bad a "Pringles can" antenna might be.)
A 10' dish has a capture area of about 78 sq ft, or 7.3 m^2. A 12'
dish has 113 sq ft, or 10.5 m^2. Dishes aren't 100% efficient, no
antenna is, but lets assume, for the sake of argument that the 10'
dish generates 35dBi, and the 12' dish approximately 40 dBi.
Now lets assume that the cards really were running below their 300mW
rating. Perhaps 24 dBm (250 mW), rather, 24.67 dBm (292.8 mW).
And, lets assume that the world is perfect, and we have ZERO losses
due to connectors and cables. ZERO.
LOS path loss at 125 miles is 146 dB. Due to reciprocity, it
doesn't matter which antenna gets the transmitter, but lets say that
we transmit
through the 40 dBi antenna. So, 24 dBm into 40 dBi for 64 dBm EIRP
(2511.89 Watts). We encounter a *MINIMUM* of 146 dB of path loss,
so the signal can't enter the remote antenna any higher than 64 -
146, or -82 dBm. The receiving antenna gets to add its gain, which
is 35dBi, so we're at -47 dBm.
This is a LONG WAY from -37dBm.
Note that if they turned the power down, the received signal level
would... go down. If the path loss was higher (it can't be lower!),
the signal level would go down.
The only possibilities left that I can account for:
1) my rough antenna gain calculations are wrong
2) they're lying, or their power meter wasn't calibrated, or ...
something.
> What on earth do they need an amp for??
> Violation of FCC part 97 rules? ---> "When the amp was used the signal
> strength bumped up to -18 dBm at 124.9 miles!"
> 97.313 Transmitter power standards.
> (a) An amateur station must use the minimum transmitter power
> necessary
> to carry out the desired communications.
> also interesting...
> "the base station end had FAA 2.4 ghz communications so we chose to
> keep our ERP low on that end"
> Why did they need an unbalanced link??
So as not to interfere with the FAA. I wonder what the back lobe on
the other side looks like. :-)
Further issues
97.307 (b) Emissions resulting from modulation must be confined to
the band or segment available to the control operator. Emissions
outside the necessary bandwidth must not cause splatter or keyclick
interference to operations on adjacent frequencies
What this says is "no out of band emissions" (from modulation, rather
than spurs). This *IS* the problem I keep pointing out.
Also, there would be a problem here, except they're under 1W of tx
power.
97.311 (d) The transmitter power must not exceed 100 W under any
circumstances. If more than 1 W is used, automatic transmitter
control shall limit output power to that which is required for the
communication. This shall be determined by the use of the ratio,
measured at the receiver, of the received energy per user data bit
(Eb) to the sum of the received power spectral densities of noise
(N0) and co-channel interference (I0). Average transmitter power over
1 W shall be automatically adjusted to maintain an Eb/(N0 + I0) ratio
of no more than 23 dB at the intended receiver.
By their own admission, the Eb/No is > 23 dB. Quoting:
> "The connection had an astounding signal strength of -37 dBm,
> bottoming out at around -50 dBm throughout the 3 hours. The noise
> levels were around -84 dBm"
> Jim,
> To be perfectly fair about it, I think that DefCon should amend
> their rules for the shootout to limit it under FCC part 15 regs and
> any stock
> Wi-Fi card. Without this restriction, the DefCon shootout is
> meaningless.
The rules should probably state that all equipment, and the resulting
system must be demonstrably able to pass the FCC Part 15
requirements, or the participants should get a special operating
waiver from the FCC. Heh heh.
They should probably also make the contest more difficult by
dividing by the measured area of the antenna(s), or at least the
stated gain.
Perhaps DEFCON should change the rules so you have to communicate
with two stations at the proposed range with a minimum angular
separation of 30 degrees.
Then it might become a bit more about engineering, and
experimentation, and less about big dishes and external amps.
> I don't have any problem with enhancing the drivers do deal with
> the distance,
for a prism-based card, its more difficult than you think. (And
really, I am aware of iwconfig <interface> txpower NN.) "Full" and
"off" work well, but these are the only positions where you have any
accuracy or repeatability on a Prism2/2.5/3.0 card.
For an Atheros card, its straight-forward. (Thoughput goes to hell
anyway.)
> but it will just become a my dish is bigger than your dish contest.
Oh, size isn't everything. I considered taking a pair of the Vivato
"prototypes" to the first contest. These *were* FCC-legal, and
*did* have 27dBi arrays. If anything they would be easier to aim,
and there is a bit of array gain as well.
Yes, I know some random Vivato distributor took their first- generation (production) Vivato "WiFi Switch" to a shootout and failed
miserably. Anyone who's read through my weblog knows why.
Frankly, if I wanted to win, and didn't care about FCC rules, I'd
just go purchase 2 of the current-generation Vivato units. These
can (if you know what you're doing) generate 30dBm (and the tx power
is adjustable in 0.5 dBm increments) into a 23dBi array, but they've
got **excellent** receiver sensitivity. (About 9 dBm better than
the rated spec on the cards these guys used.)
This means I'm fighting their 35dBi antenna with an effective 32dBi
(23dBi antenna, and 9dBm better receiver sensitivity). Yes, I give
up 3dB to their 10' dish, but my unit is 1/2 m x 1/2 m, and it won't
rattle apart while I'm driving around the desert. (Did I mention
that I'm a Las Vegas native, *and* I have a LandCruiser
(www.smallworks.com/~jim/LandCruiser) under construction in Las
Vegas, *and* I probably understand the insides of Vivato's products
as well as anyone?)
This way I could:
a) run under Part 15 or Part 97 at will. (Perhaps DEFCON will have a
class for each.)
b) actually tune the transmit power down to the minimum required to
maintain the link. (DEFCON should buy a power meter.)
c) use less power during transmit, due to a superior receiver structure.
d) very likely win, even if they cheat.
In a perfect world (true line of sight, very low noise and
interference, and high-enough mounting points), assuming that I stay
inside the FCC limits, I can generate 24dBm into that 23dBi array,
have a 23dBi array at the other end, and can decode 11Mbps down to
-92 dBm. So, in theory, I can tolerate 162 dB of path loss. On
paper, I can do 800 miles. <---- I am not proposing or promising
this in the real world.
But I'd have plenty of fade margin available on their 125 mile link. And I'd be Part 15 legal.
> ...I wonder if doppler is too much to make an EME contact using Wi-Fi?
maybe, (I don't want to have to calculate the coherence frequency for
CCK right now), but Just the path loss is 211 dB, never mind the
losses due to bouncing off the moon, and the (approximate) 2 second
round trip means there is no way you're using "Wi-Fi" (the 802.11
MAC) to close the link. You might use 802.11 gear, and some huge
antennas, but it won't be running the Wi-Fi / 802.11 MAC.
> -----Original Message-----
> On Behalf Of WA4OSH
> Why did they not open a Telnet session that is open for every observer
> to see? Why did they need an SSH tunnel on this link?
My guess was it was a habit to use SSH and it was an oversight.
> What on earth do they need an amp for??
No amp was used in the 125 mile link. They tested the amps for about 10
minutes when the remote group was 143 miles away and this link did not work.
Unfortunately the terrian would not cooperate. They had the amp on loan from
a vendor.
> Violation of FCC part 97 rules? ---> "When the amp was used the signal
> strength bumped up to -18 dBm at 124.9
> miles!"
Like I said above, no amp was used in the 124.9 mile link.
With the way the DEFCON contest seems to be run ...
You can set up something to use 12 foot dishes. Run under part 97 use a BelAir 100 (27.5 dBm and good receiver sensitivity) get them to sponsor you and modify their phy and mac slightly to accomodate DX contacts (wink-wink)
for example, increase the chipping rate and decrease the raw data rate so that you get much more processing gain...
Maybe even get the first EME contact on Wi-Fi
I really think that DEFCON needs to revisit their contest rules
> With the way the DEFCON contest seems to be run ...
> You can set up something to use 12 foot dishes.
> Run under part 97
> use a BelAir 100 (27.5 dBm and good receiver sensitivity)
oh sure.
> get them to sponsor you
unlikely, at least while I still know the VP of marketing. :-)
> and modify their phy and mac slightly to accomodate DX contacts
and... ya lost me.
1) you're not going to modify their PHY, its just an Intersil part. <--- several million dollars to develop a new part
2) even ***BELAIR*** can't modify the MAC. They paid Nesus to do
so, and while I won't reveal the figure, it was *expensive*.
might as well drop a DSP and a couple FPGAs on a board and develop
your OWN radio.
> (wink-wink)
nudge nudge
> for example, increase the chipping rate and decrease the raw data rate
> so that you get much more processing gain...
uh... use one of the modern Atheros chipsets that does this with OFDM
(course, then you can't run Part 97, but apparently rules
are made to be broken.)
> Maybe even get the first EME contact on Wi-Fi
unlikely
> I really think that DEFCON needs to revisit their contest rules
> If they keep the same rules ... I think I would like to try it next
> summer.
> Part 97 .. homebrew gear on both ends
naw, I'll call Vivato and ask if we can buy (cheaply) a pair of the
original prototypes. Then we can put some 400mW miniPCI cards
in and attach the phased array (29 dBi).
to be sure, we'll want to run Part 97.
> Does it have to run in the 13cm / 2400 Mhz band?
