Forum

I attended a Xirrus seminar today and a slide was project with this best practice for RF:
"Design coverage for signal strengths of at least
-70dBm or better"

Isn't -70dBm a very weak signal? I thought that the Receive sensitivity of many wifi devices was around that, so shouldn't a good practice be to design for a lot more? Wouldn't -70dBm mean very very slow speeds?

thanks!
Alessandra
(-10 days to the CWNA test!)

-70dBm is pretty good ..

I believe you have some concept wrong. Receive sensitivity for wireless adapters cannot be so high like -70 .. It is usually somewhere around -93dBm to -100dBm.

If you use something like NetStumbler and check your SNR, you'll see that -70dBm is actually very good.

As you go down to around -90dBm, that is when it starts to get really weak as your radio cannot differentiate between the signal and noise.

The above is from my limited knowledge and I'm no expert. So yea, don't take my word for it and wait for the experts to comment :P

Below is what I've quoted from the Official Study guide, Chapter 2 , table 2.6 showing the receive sensitivity scale for a Cisco Aironet 802.11a/b/g CardBus adapter.

−94 dBm
1 Mbps

−93 dBm
2 Mbps

−92 dBm
5.5 Mbps

−86 dBm
6 Mbps

−86 dBm
9 Mbps

−90 dBm
11 Mbps

−86 dBm
12 Mbps

−86 dBm
18 Mbps

−84 dBm
24 Mbps

−80 dBm
36 Mbps

−75 dBm
48 Mbps

−71 dBm
54 Mbps

So from the above you can see that even at -94dBm, the connectivity is "acceptable".

Vivek

Unfortunately, the number provided by manufacturers are virtually worthless in real life implementations.

Keep in mind, signal is only half the battle when it comes to communication. The other is noise.

Signal vs. noise is what really drives data rate, not just signal. When manufacturers test their devices, they are doing it in a Faraday cage which is a noiseless environment. So in the Faraday cage, they can achieve 1Mbps data rates at some crazy figure like -94dBm. In reality, you will never get a connection at such a low signal strength because there is (most) always competing noise.

-70dBm is really good signal. I'm really not happy with anything less than -79dBm in a coverage only environment. In application style and VoIP environments the minimum signal is much higher, somewhere between -70dBm and -67dBM depending on who you talk to.

Good luck on the exam!

GT

thank you.