wifigi

Hello NS Forums!

I've stumbled a few times before but was never really serious about it or had the equiptment to do so properly. Over time I acquired several antennae and cards so I thought I'd get one crazy rig together.

The Line up:
Stumbling Laptop:
-HP 600mhz PIII running XP Pro and Netstumbler 0.4.0 running marius's MapPoint + script with MapPoint 2004
-EnGenius 2011CD+EXT2 100mW card with a HyperLink Tech 14db Omni Antenna
-Linksys WUSB11 15mW usb .11b device
-Built-in 10/100 ethernet with static IP (192.168.0.1)

Remote Laptop:
-Compaq 1.3ghz Centrino running XP Pro
-Built-in 10/100 ethernet with static IP (192.168.0.2)

Other stuff
-750watt DC-AC inverter
-Magellan Meridian GPS
-'98 Chevy Blazer

The Problem
Having a very partial knowledge of radiation lobes and distribution patterns, I knew one large antenna wouldn't cut it for local networks and a small one wouldn't get those houses recessed far off the road so I said "why not both?"

The first problem I saw in doing this was that my pigtail for the HLT antenna is merely 19 inches so the laptop would have to be very close to the antenna and the antenna would be on the roof... Second, since I am running two instances of NS, only one would have a GPS lock and the other would be locked out of the COM port. Finally, if I was FORCED to have the laptop on the roof, how would I see the beautiful real time mapping and hear the speech engine at work?

The Solution
So, on the roof is the stumbling laptop. To it are directly connected the PC Card 2011CD with HLT antenna on one side of the truck and the WUSB11 on the other side. I have a modified mast mount for the HLT and use a ziptie to hold the small USB adapter. The GPS serial cable runs from the dash-mounted GPS, out the rear window, and to the laptop. Out that same window are the power cable (from inverter) and ethernet cross over cable (to remote laptop). To protect the roof laptop, it is in a padded case that is wedged between two adjustable bike rack holders and is additionally held with an elastic rope.

The remote laptop is with me in the passenger's seat. The only things that connect to this is its power cable and ethernet connection. Each laptop has a static IP so I can remotely access the stumbling laptop from within the truck via Terminal Services Client. This allows me to see, hear, and control the roof laptop without actually being up there with it. (although that could be fun)

Finally, the COM port issue. I had the aforementioned rig working for about two days before I noticed a large differentiation between the two antennae. While the HLT could get a signal from much greater distances than the WUSB11, when I was driving down the roads of tightly packed neighborhoods, much to my chagrin the WUSB11 significantly out performed the HLT. This might be due to the wider lobes of the HLT and relatively narrow ones of the WUSB11 but I'm not sure. This was all fine and dandy BUT recall that only one instance of NS received GPS coordinates. I wrote a really crappy program in C that would estimate the position of an AP based on time, MAC address, and coordinates from the GPS-locked summary BUT it was severely faulty and not worth spending hours working on something to disguise the problem at hand. With only 15 mins of googling, I found a program to fix this called TCP-COM. This program takes any RS232 signal and converts it to TCP (intended to be used with old register systems and data acquisition/control systems). I downloaded a trial version and had it listen to COM1 where the GPS came in and make virtual COM ports 4 and 5. The two instances of NS listened to their respective port so both had the SAME EXACT coordinates but had different antennae connected. Pretty cool.

Conclusion
Over three days with setup, I've had NO problems whatsoever (besides the random idiot at a traffic light ask me if I could hack his computer FROM RIGHT THERE ON THE HIGHWAY). I collected 1009 unique AP's and coords all with minimal effort. The setup takes about 20 mins each way so I don't feel it's too much of a burden at all. I've attached some pics and crude topology diagrams for your viewing pleasure.

Attached Images

	sm_Rig-001.jpg (78.7 KB, 541 views)
	sm_Rig-004.jpg (76.5 KB, 472 views)
	sm_Rig-005.jpg (85.9 KB, 467 views)
	Hardware-Topology.jpg (28.2 KB, 448 views)
	Software-Topology.jpg (37.4 KB, 359 views)

King_Ice_Flash

How long does this take you to set up before a trip?
Did you even strap the laptop down?

__________________
"Yeah," said a voice from under the table, "you go to pieces so fast people get hit by the shrapnel."

sparafina

What happens when it rains?

wifigi

The setup takes about 20 mins each way so I don't feel it's too much of a burden at all. If the two of us set up or tear down, it can be as quick as 10 mins. We definately have a good system going.

When it rains we don't go out. Even if I had a water proof setup, rain kills 2.4ghz so the trip would almost be a waste as all the SNR's would be inaccurate.

Techmasta

Wow that is pretty crazy.

King_Ice_Flash

That thing is only 14 db? Whoa! My truck is going to look like a marine vessel when my 17 gets here!

__________________
"Yeah," said a voice from under the table, "you go to pieces so fast people get hit by the shrapnel."

dwakeman

Very creative setup and thanks for the TCP-COM link, pretty cool.

Deltron3030

Sick setup man...

I was wondering where I could purchase those antenna's... I googled for a couple minutes and couldn't find anything.

~Del

sparafina

It just seems odd to have a notebook on the roof rack when you can run the cables inside. You can get a longer pigtail as well as a USB extension.

I'm not sure I understand the rationale if accuracy is of concern. The HLT will pick up signals further away, but your GPS only records the car's position. So your locations will be less accurate than using an omni with less gain.

Thorn

My own experience is that weather has very little effect on 2.4GHz. Having done WarDrives in everything from 90º plus clear summer days to thunderstorms to New England No'easter blizzards at -40º, the results are pretty consistant.

Some people have reported better results in rain, fog or snow. They claim that the water droplets aid multipath in some situations.

__________________
Thorn
"I'm The Doctor. I'm a Time Lord. I am from the planet Gallifrey in the constellation Kasterborous. I'm 903 years old and I am the man who is going to save your lives and all 6 billion people on the planet below... You got a problem with that?"

ve2vfd

I have to admit I do get slightly better results in light rain (like a drizzle) and very poor results in snow (we get the thick wet stuff a lot here).

I find this consistent with results I get when operating UHF and VHF ham frequencies.

Pat

Madhadder

Hummm

Time to get a longer Antt. and USB Cable, and get the Lappy off the roof..

__________________
Legends may sleep, but they never die!!!!

The Others

But this would take the craziness away from it

I'd get the longer cables, as suggested, and put linux on the laptop in the car. Then you can run it with both cards at the same time and leave one of the laptops at home. That way it definatly wont get rained on

__________________
all good ends all

Metaboe

when stumbling with more than one device you could get a lot of fake AP's

also when using 2 or more laptops ....

wrzwaldo

I suppose if they were operating in Ad-Hoc mode. I have never had a problem with fake AP's with multiple laptops.