Tag Archives: RSSI

More about wireless analysis

Posted on by
Reply

To expand further on the accountability aspect of the previous post, I put this little example for you all…

The first two images below are the UDP Iperf bandwidth test for a point-to-point connection and the RSSI and Link Quality assessment, using the radio’s own interface. These tests were performed outdoors, during the day (when there’s real traffic) with two 21dBi OFDM directional antennas. These tests did not involve streaming any video and were simply to test the spectrum and radios in the outdoor environment.

The first UDP Iperf bandwidth test was for a 10Mhz channel at the beginning of the frequency spectrum. We can see that it’s transmitting (during that moment) 11 Mbps and receiving 8 Mbps. The second image shows that radio one has a wonderful RSSI of -48 with a 100% Link Quality. The marketing literature states up to 17 Mbps bandwidth capabilities using a 10 Mhz channel.  So, okay. Something is not right – even though you’re thinking “Wow, 11 Mbps! I can stream like five cameras through that virtual pipe!”

The next two images show the UDP Iperf bandwidth test for another 10Mhz channel at the opposite end of the frequency spectrum. We can see that it’s transmitting (during that moment) 17 Mbps and receiving 17 Mbps, matching what the marketing materials say.

The next two images show the UDP Iperf bandwidth test for a larger 20 Mhz channel at the beginning of the frequency spectrum. We can see that it’s transmitting (during that moment) 12 Mbps and receiving 18 Mbps, which is odd, because the marketing materials state up to 35 Mbps. The RSSI shows as -48 with a 100% Link Quality, so what’s up with that? (Hint: Never design a wireless public safety architecture on RSSI alone)

The next two images show the UDP Iperf bandwidth test for a larger 20 Mhz channel at the opposite end of the frequency spectrum. We can see that it’s transmitting (during that moment) almost 35 Mbps and receiving 35 Mbps, matching the marketing materials. The RSSI shows as -47 with a 100% Link Quality.

So, what does this test tell you?  Don’t just base your architecture on the RSSI alone, as although it appears consistent and exceptional, the bandwidth tests show otherwise. Besides the fact that this test was a single city block or two away from each other so the RSSI should’ve been closer to a burning -30 or less RSSI. Also, there’s something odd going on within the first half of the frequency spectrum to cut the bandwidth in half.

The spectrum analysis of the area showed that there was something using the first half of the spectrum, but not necessarily using it at that time. The image below shows a device using “auto channel select” – scrolling up and down the first half of the frequency looking for a mobile vehicle to link up with. Probably why it didn’t affect the RSSI at that very moment because whatever vehicle and/or station didn’t connect, it was just looking.

So there you go…

Smart antennas aren’t that smart

Posted on by
Reply


I’ve received different instructions regarding the proper cabling of these new smart MIMO antennas. MIMO (multiple input, multiple output) is the new smart antenna technology for 802.11n wireless communications in which multiple antennas are used at both the transmitter and the receiver. These smart antennas are the reason we see bandwidth of 150Mbps + on 40Mhz channels.

I’ve created a diagram that shows that crossing cables 2 and 3, in a point to point configuration, can be beneficial, yet when doing this in the field, the RSSI did not reflect any change because I don’t believe these radios and their smart antennas have reach maturity. The biggest issue I’ve found with using the Firetide 7000 series MIMO radios, with their 19dBi 20 degree directional MIMO antennas is their back-lobes. Clearly, these back-lobes aren’t shielded enough as when working on linear meshes, these antennas cannot be back-to-back on the same pole – at all.

First of all, the RSSI is calculated based on connection #1 only, so testing bandwidth is the only way to confirm that the crossing of connection #2 and connection #3 would create any improvement. I’ve seen kinked cables on #2 & #3, without any effect on RSSI, but clearly visible in bandwidth tests.

When creating a point-to-point connection, front-to-front, the vertical polarization crosses the two angled polarization, and so it would be better to flip one antenna 180 degress (see diagram). Easiest method of remembering the proper configuration is having one vertical polarization arrow pointing up, while it points down at the other end.

Otherwise, chaos ensues as any other antennas pointing in the same direction create an even better (unintentional) connection front-to-back (where the polarizations match exactly), in close proximity, making it a problem to distinguish between radios and antennas without visual confirmation of the radios, antennas and cable connections.

Considering how the smart antenna is supposed to work, and the new proliferation of MIMO technologies, self-interference and network loops can become an issue without thorough design considerations. Just because you can’t see the link, doesn’t mean it isn’t there.