Forum

Hi all, I have a query which may be straightforward for some forum readers to answer.

A client device uses the RSSI to work out the data rate to use (I know SNR, retransmissions, etc may also be a factor but let me keep this simple for now). OK, let's say a device records a RSSI of -67dBm and decides to use 54Mb/s. If he moves away the RSSI will get lower (maybe -79dBm) and a lower data rate will be used. That's all cool....

My question, how does the client work out the RSSI? How does it calculate the signal as being, for example, -67dBm? I have always made a VERY BIG ASSUMPTION that it is based on the amplitude of the received signal - the higher/bigger the amplitude the stronger the signal, hence better the RSSI. I haven't actually read this anywhere though and this is a fundamental cornerstone for me to understand client devices.

Can any clarify?

Thanks

DJ

Hi Darren

When we transmit a Wi-Fi signal, the input data stream and any required overhead are used to modulate a carrier signal which is generated internally. There are various types of modulation. For example, QPSK, QAM etc.

Why do we even need to modulate a carrier signal ? Why not just try to send the data straight out of the antenna ? There are a few reasons behind this. Firstly, data signals ( imagined as digital pulses coming in usually via Ethernet and then being converted to logic levels inside the Wi-Fi device ) do not radiate well from antennas. Depending on the type of wire used, we do know that when Ethernet signals are sent down cabling, that electromagnetic radiation does emanate from the cable. In order to prevent interference of one system to another, we use techniques such as shielding, wire twisting etc to minimize this radiation. We also try to separate power cabling from data cabling whenever possible, due to the 50 Hz or 60 Hz "Hum" present .

Rectangular or square digital pulses can be represented mathematically as a single "fundamental" frequency and a number of odd "harmonics" or multiples of the fundamental frequency. These are sine waves and can be thought of as radio waves. A Frenchman called Fourier discovered this, and the Fourier theory is very important in electronic/RF  engineering. Many of the UPS systems in Data Centers utilize this theorem. When we take in a "dirty" AC supply, a rectifier coverts it into DC. A device called an inverter then converts the DC to square pulses. Fourier tells us that the square wave is made up of a fundamental or root frequency plus a bunch of harmonics. A harmonic filter then removes the higher harmonics and then leaves us with a nice clean fundamental AC frequency which can power our Wi-Fi gear, servers etc.

The next reason is that the major Wi-Fi frequencies are in ISM type bands. These use much higher frequencies than the data signals ( even including the highest harmonics ). Hence, we need to "piggy back" the data signals on a "carrier frequency". That carrier frequency is in the appropriate band.

Once we do this, we have what is known as a modulated signal.

If we were to look at a "clean carrier" i.e. no modulation, on a spectrum analyzer, we would see a "spike' sticking up at the particular frequency of interest.

If we were now to look at a modulated signal, we would see ( depending on the modulation type ), a "spreading" of the energy. In other words, instead of a "spike", we would have, perhaps, a more rectangular shape.

That rectangular shape has a certain "area". In order words, if we were to break down that shape into a number of tiny slices, we could work out the area of each little slice and add them all together ( a process of integration, as in calculus ). This would help us to work out the total power of the modulated signal.

Using sophisticated filtering techniques ( and digital signal processors ), the Wi-Fi card demodulator can work out the total power of the demodulated signal.

That is then tied into the RSSI value. In other words we are measuring power and not voltage.

In a nutshell, the RSSI value is obtained from calculating the "area" of the incoming signal, and not the center frequency amplitude. However, if we were to remove the input data stream to the modulator at the other end, we would have a spike, whose amplitude would normally equal the "spread out" modulated power of the signal when the data stream was reintroduced.

In RF engineering, we do not normally like "clean carriers" being transmitted ( except when lab testing and in some controlled environments .

Dave