I'll say it's max! 802.11b/g will do distances like this with decent antennas too, but saying 30 miles and non-LOS in the same breath is horse hooey.
Let's see, assume transmitter power like WiFI, say 11a and +20 dBm (100 milliwatts). Allow signals to occupy spectrum, say 1 bps/Hz gives 70 MHz bandwidth. Assume this is broadcast, with 10 dB of gain in the vertical plane, a la cell phone cells. +30 dBm ERP, receiver noise floor is KTB=-96 dBm or so in 70 MHz (no interference, thermally limited noise floor). So we have 126 dB of C/N to play with. Now, grant another 10 dB of receiver antenna gain, (can't have too much directivity because users aren't going to be willing to constantly repoint their antennas). 136 dB of C/N. Let's say the modulation scheme can make do with 6 dB for low BER. 130 dB C/N for path loss.How long is this path? Depends on frequency but at 5 GHz it's about 37+20Log(Dmiles) + 20log(5000)=111 dB at 1 mile. So we can go 19 dB (call it 20) further or 10 miles. To get to 30, we'll have to use more than 10 dB antennas, but 15 dB antennas on both ends would do it. Yes, it can work with reasonable hardware, LOS at 30 miles, just like 802.11a or similar. BUT
NOW the fun begins. What happens when the *real world* enters with non-LOS paths? Right, it depends upon the path. If it's through a hill, another 80 dB of path loss is easy to imagine. In fact if it's even through a moderate tree, another 20 dB is easy. One tree! That dropped our 10 miles back to 1 mile. A few trees and are easily at another 20 dB of path loss and down to.1 mile. Is this beginning to sound like your experience with WiFi, "1300 feet outdoors" less indoors? For good reason! It's governed by the same laws of physics and real world paths. It will act the same, no matter what Intel says or what the protocol is designed to do. This is physics, Shannon and real world paths set the limits.
Intel may sell a lot of hardware and make more money but they *aren't* going to be capable of building radios simultaneously capable of 30 miles and non-LOS. If it were possible, your cell phone which runs more power (perhaps several watts)at frequencies where there is lower pathloss wouldn't drop your call 2 km from the cell site. And at that a call which only requires on the order of 10 kbps. 10,000 times that datrate on less power at higher frequences and 10X the distance. Let's get real.
Slashdot Mirror
I'll say it's max! 802.11b/g will do distances like this with decent antennas too, but saying 30 miles and non-LOS in the same breath is horse hooey.
.1 mile.
Let's see, assume transmitter power like WiFI, say 11a and +20 dBm (100 milliwatts). Allow signals to occupy spectrum, say 1 bps/Hz gives 70 MHz bandwidth. Assume this is broadcast, with 10 dB of gain in the vertical plane, a la cell phone cells.
+30 dBm ERP, receiver noise floor is KTB=-96 dBm or so in 70 MHz (no interference, thermally limited noise floor). So we have 126 dB of C/N to play with. Now, grant another 10 dB of receiver antenna gain, (can't have too much directivity because users aren't going to be willing to constantly repoint their antennas). 136 dB of C/N. Let's say the modulation scheme can make do with 6 dB for low BER. 130 dB C/N for path loss.How long is this path? Depends on frequency but at 5 GHz it's about 37+20Log(Dmiles) + 20log(5000)=111 dB at 1 mile. So we can go 19 dB (call it 20) further or 10 miles. To get to 30, we'll have to use more than 10 dB antennas, but 15 dB antennas on both ends would do it. Yes, it can work with reasonable hardware, LOS at 30 miles, just like 802.11a or similar. BUT
NOW the fun begins. What happens when the *real world* enters with non-LOS paths? Right, it depends upon the path. If it's through a hill, another 80 dB of path loss is easy to imagine. In fact if it's even through a moderate tree, another 20 dB is easy. One tree! That dropped our 10 miles back to 1 mile. A few trees and are easily at another 20 dB of path loss and down to
Is this beginning to sound like your experience with WiFi, "1300 feet outdoors" less indoors? For good reason! It's governed by the same laws of physics and real world paths. It will act the same, no matter what Intel says or what the protocol is designed to do. This is physics, Shannon and real world paths set the limits.
Intel may sell a lot of hardware and make more money but they *aren't* going to be capable of building radios simultaneously capable of 30 miles and non-LOS. If it were possible, your cell phone which runs more power (perhaps several watts)at frequencies where there is lower pathloss wouldn't drop your call 2 km from the cell site. And at that a call which only requires on the order of 10 kbps. 10,000 times that datrate on less power at higher frequences and 10X the distance. Let's get real.