I interviewed the team after the awards ceremony. They dialed down the 300 mw cards to 30 mw.
Humphrey Cheung Editor www.tomshardware.com -----Original Message----- From: [email protected] [mailto:[EMAIL PROTECTED] On Behalf Of Jim Thompson Sent: Monday, August 01, 2005 12:15 AM To: [email protected] Subject: [SOCALWUG] Re: Defcon WiFi Shootout Record Set at 125 Miles for 802.11b On Jul 31, 2005, at 2:25 AM, [EMAIL PROTECTED] 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: antenna gain: 6.00 dBi, tx power = 30.00 dBm (1000.00 mW), EIRP = 36.00 dBm, (3.98 W) antenna gain: 7.00 dBi, tx power = 29.67 dBm (926.12 mW), EIRP = 36.67 dBm, (4.64 W) antenna gain: 8.00 dBi, tx power = 29.33 dBm (857.70 mW), EIRP = 37.33 dBm, (5.41 W) antenna gain: 9.00 dBi, tx power = 29.00 dBm (794.33 mW), EIRP = 38.00 dBm, (6.31 W) antenna gain: 10.00 dBi, tx power = 28.67 dBm (735.64 mW), EIRP = 38.67 dBm, (7.36 W) antenna gain: 11.00 dBi, tx power = 28.33 dBm (681.29 mW), EIRP = 39.33 dBm, (8.58 W) antenna gain: 12.00 dBi, tx power = 28.00 dBm (630.96 mW), EIRP = 40.00 dBm, (10.00 W) antenna gain: 13.00 dBi, tx power = 27.67 dBm (584.34 mW), EIRP = 40.67 dBm, (11.66 W) antenna gain: 14.00 dBi, tx power = 27.33 dBm (541.17 mW), EIRP = 41.33 dBm, (13.59 W) antenna gain: 15.00 dBi, tx power = 27.00 dBm (501.19 mW), EIRP = 42.00 dBm, (15.85 W) antenna gain: 16.00 dBi, tx power = 26.67 dBm (464.16 mW), EIRP = 42.67 dBm, (18.48 W) antenna gain: 17.00 dBi, tx power = 26.33 dBm (429.87 mW), EIRP = 43.33 dBm, (21.54 W) antenna gain: 18.00 dBi, tx power = 26.00 dBm (398.11 mW), EIRP = 44.00 dBm, (25.12 W) antenna gain: 19.00 dBi, tx power = 25.67 dBm (368.69 mW), EIRP = 44.67 dBm, (29.29 W) antenna gain: 20.00 dBi, tx power = 25.33 dBm (341.45 mW), EIRP = 45.33 dBm, (34.15 W) antenna gain: 21.00 dBi, tx power = 25.00 dBm (316.23 mW), EIRP = 46.00 dBm, (39.81 W) antenna gain: 22.00 dBi, tx power = 24.67 dBm (292.86 mW), EIRP = 46.67 dBm, (46.42 W) antenna gain: 23.00 dBi, tx power = 24.33 dBm (271.23 mW), EIRP = 47.33 dBm, (54.12 W) antenna gain: 24.00 dBi, tx power = 24.00 dBm (251.19 mW), EIRP = 48.00 dBm, (63.10 W) antenna gain: 25.00 dBi, tx power = 23.67 dBm (232.63 mW), EIRP = 48.67 dBm, (73.56 W) antenna gain: 26.00 dBi, tx power = 23.33 dBm (215.44 mW), EIRP = 49.33 dBm, (85.77 W) antenna gain: 27.00 dBi, tx power = 23.00 dBm (199.53 mW), EIRP = 50.00 dBm, (100.00 W) antenna gain: 28.00 dBi, tx power = 22.67 dBm (184.78 mW), EIRP = 50.67 dBm, (116.59 W) antenna gain: 29.00 dBi, tx power = 22.33 dBm (171.13 mW), EIRP = 51.33 dBm, (135.94 W) antenna gain: 30.00 dBi, tx power = 22.00 dBm (158.49 mW), EIRP = 52.00 dBm, (158.49 W) antenna gain: 31.00 dBi, tx power = 21.67 dBm (146.78 mW), EIRP = 52.67 dBm, (184.79 W) antenna gain: 32.00 dBi, tx power = 21.33 dBm (135.94 mW), EIRP = 53.33 dBm, (215.44 W) antenna gain: 33.00 dBi, tx power = 21.00 dBm (125.89 mW), EIRP = 54.00 dBm, (251.19 W) antenna gain: 34.00 dBi, tx power = 20.67 dBm (116.59 mW), EIRP = 54.67 dBm, (292.86 W) antenna gain: 35.00 dBi, tx power = 20.33 dBm (107.98 mW), EIRP = 55.33 dBm, (341.45 W) antenna gain: 36.00 dBi, tx power = 20.00 dBm (100.00 mW), EIRP = 56.00 dBm, (398.11 W) 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. Their website (http://www.wifiworldrecord.com/) says they borrowed amps, btw. Perhaps they didn't use them? 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
