I am trying to connect to a wireless service provider that is about 5 miles from me, as the crow flies. I can't directly see their antenna but am just on the other side of a slope from them. What would be the best solution to getting connected to them? BTW- I am living in the Balkans and have little immeadiate availability to equipment. What I do have is a DLink 900-AP+ and a DLink ANT24-1400 outdoor antenna. Can anyone let me know if these things will work or what some good ideas are for improvising?
[POSTED TO alt.internet.wireless - REPLY ON USENET PLEASE]
The "2.4GHz Antenna Gain vs. Range Table" at
shows the range difference (under ideal conditions) between 14 dBi and 24 dBi antenna gain, and suggests that 14 dBi *might* be able to work at low speed over your 5 mile (8 km) distance (assuming a good omnidirectional antenna at the other end).
No. At 5 miles, even with line of sight, and a 24dBi dish antenna, and mounting your client radio next to the dish, you still will not have enough signal to make it to the truck stop because they're probably using a simple wireless router with the stock omni directional antennas. They would need to either install an outside antenna for your benefit, or possibly install an additional access point with directional antenna specifically for your benefit. I calculate the fade margin if you want, but I can speculate that it's not going to happen even with line of sight.
Incidentally, at 2.4GHz, you need *MORE* than just line of sight. You need Fresnel zone clearance.
5 miles, you need a radius of 52ft of clearance at the midpoint. That means that the center line connecting your antenna and the one at the truck stop has be at least 52 ft off the ground at midpoint. Another combination is if you assume the truck stop end to be at ground level, your antenna needs to be 104 ft off the ground.
Also, if there are any other 802.11b/g stations along the line of sight (or beyond), your antenna is going to hear them quite well, and you'll find yourself dealing with interference problems.
I've seen things like that but they're often not reliable. I could hear a signal at my that I could never get a reliable direction to identify. More strangely, I could see it with one of my rooftop antennas, but none of the other two. It was an open access point, so I just connected, ran traceroute, found the ISP, and eventually identified the customer. It was literally on the other side of a hill about 4000ft away. No line of sight and many trees. The signal disappeared completely the day a nearby neighbor removed their aluminium storage building and replaced it with one that was less aesthetically disgusting. No signal at all, zero, nothing. Looking at the building, I noticed that it's large windows formed a perfect entrance way, while the inside formed a perfect 3 corner reflector. The walls were corregated but apparently that didn't make any difference. The building also had line of sight to my house and the owner of the open access point. I'm still amazed at how well it worked.
Most of the time, just miraculous propogation exercises are anything but stable and will vary considerably with changes in the local RF environment. I installed a point to point wireless link between two buildings for a customer. The antennas were 19dBi dishes and were about 50 and 30 ft off the ground respectively. Yet, if anyone parked in a specific parking space that was roughly midway between the two antennas, the signal would waver and often fade to zero. The reflection off the vehicle would cancel with the direct path and cause signal cancellation. One antenna was eventually moved (by someone else) and the phenomenon was moved to the middle of the parking lot. Now, only when the UPS truck parks there, is there a problem.
Chances are very high that even if you do get a NLOS (non-line-o-sight) signal, it will not stay put as things in between move around.
Well, yeah. It would be interesting to try. I would predict disaster, but I guess it's more helpful to run the numbers. Midnight, and I should be doing something else, but I'm curious to see how close my seat of the pants guesswork hits. Let's run the numbers.
The following was lifted from the DI-624 datasheet but is close enough for most similar 802.11g radios: * 54Mbps OFDM, 10% PER, -68dBm) * 48Mbps OFDM, 10% PER, -68dBm) * 36Mbps OFDM, 10% PER, -75dBm) * 24Mbps OFDM, 10% PER, -79dBm) * 18Mbps OFDM, 10% PER, -82dBm) * 12Mbps OFDM, 10% PER, -84dBm) * 11Mbps CCK, 8% PER, -82dBm) * 9Mbps OFDM, 10% PER, -87dBm) * 6Mbps OFDM, 10% PER, -88dBm) * 5.5Mbps CCK, 8% PER, -85dBm) * 2Mbps QPSK, 8% PER, -86dBm) * 1Mbps BPSK, 8% PER, -89dBm)
My guess is that the truck stop only does 802.11b which means that you're going to be hurting for sensitivity. So does the DWL-900AP+. That means 1-2Mbits/sec or -89dBm receiver sensitivity at best.
The DWL-900AP+ belches about +15dBm. I'll assume that truck stop has something like a DI-624 or WRT54G which transmits at about +17dBm. However, I'll also assume that they're using the stock rubber ducky antennas which have an optimistic gain of about +2dBi. I'll assume you do the right thing and buy the biggest practical 24dBi barbeque grill dish.
We're looking for a fade margin of at least 20dBi. Anything less than that will be flakey and unreliable. Anything under 10dBi fade margin literally won't work. Incidentally, It is possible to correleate fade margin with system reliability and link uptime. Details when I'm awake.
Starting at your end Tx power = 15dBm Tx cable loss = 3dB (including connectors) Tx ant gain = 24dBi Distance = 5 miles Rx ant gain = 2dBi RX cable loss = 0dB (directly attached) RX sens = -89dBm (for 54mbits/sec)
Plugging the numbers into:
an operating margin of -0.2dB with a receive signal level 0.2dB below the rx sensitivity reference level. To make this link work, you'll need to pickup 20dB of gain from somewhere. If the truck stop installs a 14dBi omni on their roof for your benifit, you'll still be
6dB short of the necessary fade margin. If they install a 24dBi dish (less some connector and coax losses) pointed directly at your house, it will work, but I don't think they'll do that.
So, go ahead and try it, but I'll stand my prediction of failure.
Don't fall off the roof or tower. Otherwise, it doesn't hurt to try it and see what happens. You might be entertained by putting the entire affair in a vehicle and trying it initially from very close to the truck stop. Then move farther and farther away. If there are some nearby hills with line of sight, try those. You'll get a feel for how well it will work from such experiments. As for my motto, it really means you really haven't learned anything about wireless until you've made all the traditional mistakes and then understood why they don't work. Good luck and may your RF connectors be free of rain water.
If you have problems selecting the appropriate god or religion, perhaps the following might marginally help:
The wireless gods also accept donations. Just send the donations to my PayPal address below. However, we offer no guarantees of miraculous performance, distance, reliability, or interference mitigation.
The signal can often bend a lot more than most people realise or appreciate. Locally there is an AP some 3Km away and obscured by a decent amount of earth and bush yet I consistently get very reliable link to it.
The best thing to do is to try it and see if you can get sufficient signal with the equipment that you have but do not expect to get 54G performance. You may be able to get a 1 or 2 Mbs rate successfully though. You may need to try moving the antenna around to slightly different locations and need to aim it quite accurately.
Try it, it may just work, if not, you will probably get a better idea of what you need to do. To Paraphrase Jeff L., "Learn by Destroying" except that you will not be destroying anything.