RSSI=Received Signal Strength Indication which generally appears as "bars" in
our software. It indicates how much power is in the radio signal your router/AP
is sending you. The RSSI number also determines whether CTS (clear to send) is
set to on or off. An RSSI of 0 means no signal and CTS if set to off. RTS of 1
or above generally set CTS to on but that is very poor signal strength. The RSSI
number can go as high as 255(?) but different companies use different numbers.
Cisco goes from 0-100, Netgear, I think, only goes up to 60 for a top signal
strength. It varies.
RSSI is a sampled number so I‘d say it‘s not a 100%
accurate figure, just my opinion, more of a close
approximation.
Netstumbler is simply one (or the) most popular wireless
network monitoring tools. There are other software packages but RSSI is pretty
much measured the same way. Netstumbler is probably the most widely used
freeware product. I‘ve never bothered with any other program to compare it
to.
RSSI is being replaced by RCPI (received channel power indicator).
You might want to consider dropping your research on RSSI and instead switch to
RCPI or combine both in your research.
Addition: the RSSI is part of the preamble in the signal from a router/AP. That‘s how a client gets and measures the RSSI or RCPI. This makes sense since if the RSSI is zero, you aren‘t going to get a preamble anyway. If the RSSI is 1 or above, the client needs to know some data about how much to amplify the signal. RSSI is read before signal amplification, another reason why I think it‘s more of an approximation rather than an exact number.
RSSI is estimated, in consumer WiFi as either a marketing-speak percentage or
as a received power level in dBm. On the latter (dBm), the numbers are very
approximate and are not actually calibrated to the standard where 0dBm =
1mWatt.
Most WiFi products kind of "cheat" by estimating power based on
detected energy after demodulation. The (baseband) signal level, in the absence
of a received signal, is approximately the noise floor of the receiver. For a
certain receiver bandwidth (design parameter) the noise power can be a constant,
to which one adds the noise figure of the receiver. The sum is roughly the noise
floor - if there is no signal present, to include non-WiFi and adjacent
channels.
So the reported RSSI is derived from the noise floor in ratio
to a different time when there is a WiFi signal present. But there may be two or
more signals present, and adjacent channel power present.
So this scheme
for reporting dBm RSSI is just an approximation- good enough for its purpose, if
there is no strong adjacent or same channel interference. WiFi radios may make
this measurement during one of the low speed bits in the frame preamble. This is
also when the radio may decide to try the other (switched diversity)
antenna.
原文：http://www.cnblogs.com/sunshinePlus/p/3759897.html