Wi-fi

Have you thought about replaceing your mini pci card with a better card?

Omni-directional

Sure your card may be junk. If that does not cure it, search for an article called:

"What is the best Wi-Fi antenna for me?"

That should cover what you want to know about Wi-Fi range.

I am currently connected with a network 2.2 miles across the bay, using a Hawking HWU54D adapter:

BTW, it outperforms the newer 8 dbi model.

Dave Rudisill hath wroth:

Amazing. Let's play with the numbers. The Hawking HWU54D does not specify an antenna gain or power output. I'll assume it's at least

8dBi gain, about 1 dB in internal connector and coax losses, and perhaps +17dBm power output (optimisticly). I have no idea what you're connecting to across the bay, but lets see how big an antenna they'll require. I'll assume 802.11b at -84dBm receive sensitivity. To get any kind of reliable connection, you need an absolute minimum of about 13dB of fade margin. I prefer 20dB but 13dB will barely work. Plugging: TX power = +17dBm TX coax loss = 1dB (connector loss) TX ant gain = 8dBi Distance = 2.2 miles RX ant gain = unknown RX coax loss = 3dB (coax and connectors near antenna) RX sens = -84dBm (at 5.5Mbits/sec) Fade margin = 13dB Plugging into: trying different receive antenna gains, I get 20dBi for the access point antenna. That will need to be a fairly large sector antenna pointed directly at you. It just might work. If I went for my usual 20dB fade margin, the 27dBi gain antenna at the access point would be impossibly large.

However, we're not done yet. The antennas on both sides might be fairly close to the water level causing the water to impact the Fresnel zone. At 2.2 miles, the Fresnel zone at midpoint is a 35ft radius along the line of sight. That means that the antennas at both ends will need to be at least perhaps 30ft above the high water line for this path to work reliably.

It's VERY close to non-functional, but if everything were perfect, it should play. So, how well does it work over the water? I usually have a hell of time dealing with fog attenuation, multipath reflections off the water surface, and inversion layer refraction.

Actually, the HWU54D is only 6 dBi.(!)

According to our GPS, The Hawking is approximately 20' above sea level. Through binoculars, I can see the park with the network I am using, but when I drove over there last month I couldn't identify their antenna. They are right on the beach, as we are, but their antenna is probably somewhat higher than we are. I drove to the point in the park closest to where we are parked, and the GPS said it was 2.2 miles away.

I also have a Hawking HWU8DD adapter, which has an 8 dbi antenna. The antenna on this adapter is much more directional than the one on the HWU54D, making it much more difficult to find a distant network. Once it is zeroed in on a network, Net Stumbler and real use shows that it performs no better than the HWU54D.

Currently, Net Stumbler shows I have a SNR varying between 10 and 16. With this adapter, that's enough to provide a reliable connection. Sometimes the signal disappears for hours at a time, then suddenly comes back. Atmospherics, I presume.

Dave Rudisill hath wroth:

Yuck. I couldn't find any numbers.

GPS is a terrible way to do elevation. It's not really that accurate near the ground. If you're in Alaska, try a USGS map or find thyself on:

elevation is above mean sea level. Figure out the high tide in your area and use that as the base level for the Fresnel zone calculations:

I have fun trying to find hidden wireless antennas. I have one installation where the access point and antenna are hidden inside a plastic owl.

That's the effect you'll get when you operate near the receiver threshold. Methinks you will need a bigger antenna to make this link reliable.

Yech. If that's the difference between the signal and the noise levels displayed by Netstumbler, you've got a really marginal link. It should be at least 20dB. My tinkering seems to show I need about 35dB for a really reliable connection.

I dunno. From your description, it's not that reliable.

Nope, not atmospherics. It's reflections off the water. The direct path signal (incident) cancels with the reflected signal (reflected) and creates a null.

Here's a fun experiment. Place the antenna where it's not going to move or be affected by moving objects (i.e. furniture or cars). Build a graph of several days SNR versus time of day. Then, compare the results with the tide tables. This works best with flat water. You'll find that at some elevation of the tide, your signal will just disappear. When the tide goes up or down, it will come back. With large variations in tide height, you may get more than one cancellation point per tide cycle. If you're unfortunate enough to have cancellation at high tide or low tide, then you'll have hour long outages. I've done this with links across Monterey Bay, CA and found almost perfect correlation between fades and the tides.

If you have inversion layer problems (unlikely in Alaska) then you can also enjoy the wonders of atmospheric refraction. Lot of fun in Smog Angeles.

Actually, we are camping on a beach in Mexico. Since the beach is a few yards away, I would say we are pretty close to sea level. Topozone doesn't include Mexico anyway.

That is consistent with my experience, but I very often find much weaker signals that are good enough for getting my email and Usenet messages. With the Hawking adapter, I an usually stay connected with a signal in the 15 dB range. WWW use is pretty much out of the question with a link that weak, but it works for email.

"Reliable" is relative when you are scrounging for a signal, with no other source of Internet access. I am very pleased that most of the time I can get on the 'Net at all.

That must be it. The signal will be absolutely gone for a couple hours, then come back "strong" for a few hours. Right now, it's close to high tide, and I have an excellent connection, as I did 12 hours ago (the last high tide). A few hours before that, the network was undetectable. It has been that way for weeks. I never thought to associate it with the tides.