I agree with coder27. There is number of antennas and number of streams. You can have multiple antennas on the TX and/or RX side, and thereby increase the link budget. One way to look at this is if you have two antennas xmitting 10mW, you're xmitting 20mW. If you have two antennas, you have twice the amt of RX energy you're capturing. You actually get even more improvement than 2x (i.e. 3dB) because MIMO takes advantage of the multipath reflections. If it's not a rich multipath environment with a delay spread less than the symbol period, you'll just get the 2x improvement from doubling the antennas, which is pretty cool.
If you have good signal strength and a nice multipath channel, you can do dual streams which doubles throughput. Theoretically you do more than two streams. I have seen dual stream work on an 802.11(n) card in my own testing.
I've experimented with 802.11(n) 2-stream and seen it work. Here are the caveats:
Two get double the throughput, things have to be just right. It helps if antennas a different polarizations and/or separated.
Sometimes you're better off just using a 40MHz (or 80MHz channel) and getting your extra throughput that way.
Most commercial equipment has a poor algorithm to select when it's a good time to use dual-stream.
If the data is going over TCP, delays waiting for the TCP ACK to works its way through become a significant impediment. Doubling underlying data rate doesn't mean twice the user data rate.
Despite all this, you really can do the seeming impossible: You can double the amt of data you send over a 20MHz channel simply by adding another antenna and doing dual stream. You're transmitting two separate streams on the same frequency at the same time, which seems like magic. You can read more about 802.11(n) and 802.11(ac) MIMO here.
I agree with coder27. There is number of antennas and number of streams. You can have multiple antennas on the TX and/or RX side, and thereby increase the link budget. One way to look at this is if you have two antennas xmitting 10mW, you're xmitting 20mW. If you have two antennas, you have twice the amt of RX energy you're capturing. You actually get even more improvement than 2x (i.e. 3dB) because MIMO takes advantage of the multipath reflections. If it's not a rich multipath environment with a delay spread less than the symbol period, you'll just get the 2x improvement from doubling the antennas, which is pretty cool.
If you have good signal strength and a nice multipath channel, you can do dual streams which doubles throughput. Theoretically you do more than two streams. I have seen dual stream work on an 802.11(n) card in my own testing.
I've experimented with 802.11(n) 2-stream and seen it work. Here are the caveats:
Two get double the throughput, things have to be just right. It helps if antennas a different polarizations and/or separated.
Sometimes you're better off just using a 40MHz (or 80MHz channel) and getting your extra throughput that way.
Most commercial equipment has a poor algorithm to select when it's a good time to use dual-stream.
If the data is going over TCP, delays waiting for the TCP ACK to works its way through become a significant impediment. Doubling underlying data rate doesn't mean twice the user data rate.
Despite all this, you really can do the seeming impossible: You can double the amt of data you send over a 20MHz channel simply by adding another antenna and doing dual stream. You're transmitting two separate streams on the same frequency at the same time, which seems like magic. You can read more about 802.11(n) and 802.11(ac) MIMO here.
