2
$\begingroup$

My understanding of an RSSI reading is that it should equal Tx power - Tx feed loss + Tx antenna gain - pathloss + Rx antenna gain - Rx feed loss. If this is the case, if the radios at either end are identical and have the same Tx power, shouldn't the RSSI experienced at both sides be the same?

I have a scenario where on one side, the radio is receiving at around -100dB, at at the other side, the RSSI is -85dB. What could be the cause of this?

I have noted that both of the antennas could benefit from being raised away from the nearby metal roof, but wouldn't that still factor the same in both transmitting and receiving? Also, note that on one end there is an omni directional antenna and on the other end a yagi (same vertical polarisation). Unfortunately I do not have VSWR readings of either side at this stage.

Edit:

@tomnexus, your answer is insightful. There is some additional information which I have discovered though, which might confuse the matter even more. At this stage I cannot measure the VSWR as the site distant from here - all I have are some computer diagnostic logs which I am reviewing now, and the diagnostic utility doesn't have VSWR readings.

The radio network is in fact a point to multipoint data radio network. Let's say the base (omni antenna) is called "BASE", and the two remotes are R1 and R2, each with yagis pointing back to the base. They are not near one another. Communication is half duplex, so R1 and R2 can only hear B, not one another. BASE's RSSI from R1 is -80dBm. R1's RSSI from BASE is likewise -80dBm. BASE's RSSI from R2 is -85dBm, however R2's RSSI from BASE is -100dBm as explained above. The initial thought, seeing that R2 receives from BASE 15dB quieter than vice versa is as you said - bad VSWR at BASE, combined perhaps with the BASE radio automatically turning it's own Tx power down. However, if this were the case, I would expect to see the same phenomenon with RSSI at R1. The odd one out in this picture appears to be R2 alone. Is it ever possible that receive sensitivity is degraded more than its transmit performance? Does this point necessarily to an issue with the radio itself? As mentioned before, all antennas are each mounted rather close to the slightly angled roof of a metal shed. Is it at all possible that a radiation pattern or path loss can be influenced differently on Tx than on Rx?

I'm not sure if it's relevant or not, but the last thing is that the supply voltage at BASE is occasionally dropping slightly - this might be related or not. The nominal voltage for the radios is 13.8VDC, however these radios are commonly used in conjunction with batteries and solar radios and have a fairly wide voltage tolerance (will work properly at >10VDC, shuts down at < 9VDC). I have noted that the supply voltage for the BASE radio is 11.9, and occasionally it simply drops to 11.4 (all along, however, R2 receives at -100dBm). It might be that the power supply/battery combination is not adequately handling the required current draw, but I wonder if this current draw is increased on account of bad VSWR. That still doesn't explain it though. Thanks for you ongoing help. BTW, radios are running at roughly 5W.

$\endgroup$
2
$\begingroup$

My understanding of an RSSI reading is that it should equal Tx power - Tx feed loss + Tx antenna gain - pathloss + Rx antenna gain - Rx feed loss. If this is the case, if the radios at either end are identical and have the same Tx power, shouldn't the RSSI experienced at both sides be the same?

Your intuition about reciprocity is correct but your understanding of what RSSI means is not.

Usually, the RSSI is nothing more than a measurement of the AGC level. The AGC's job is to adjust the gain of the receiver's amplifiers so the input to the demodulator is at an appropriate level. When the received power goes up, the AGC turns down the gain, and the RSSI goes up. This prevents saturation or clipping in the demodulator, while simultaneously assuring the input level is high enough to overcome the demodulator's internal noise floor, and (for digital receivers) to effectively utilize the DAC's full range.

The AGC is just looking at the overall level of the signal coming from the antenna. It doesn't know about demodulation, so it can't distinguish between signal and noise. This is necessarily so, because if the AGC were not to reduce gain in the presence of noise the demodulator would become saturated, causing distortion and effectively more noise.

When you see an RSSI associated with a particular packet, this is usually a sample of the AGC level as that packet was being received.

So, your situation could easily be explained by a noise source which is closer to one station than the other. While the signal between the two stations obeys reciprocity and thus should have identical strength on each end, the higher noise at one station requires the AGC to reduce gain, and thus increase RSSI.

Also keep in mind that there's no defined unit for RSSI. It's not necessarily in dBm, or even decibels relative to any reference at all. So a difference of 15 in RSSI does not necessarily mean a 15 dB difference in signal strength.

$\endgroup$
  • $\begingroup$ Thanks for this. What you're saying does make sense. However, the units displayed on the diagnostic software specific for these radios is dBm for RSSI - not sure why this may be the case. So if this is what's happening - a noise source which is closer to on station than the other - I should be getting the same -85dB reading at BASE even when I disconnect R2. I'll have to do this test when I'm on site next time. $\endgroup$ – epv Apr 23 '16 at 1:16
0
$\begingroup$

VSWR is one possible explanation.

If the transmitters are a few watts or more, they probably have a protection circuit that reduces the transmitter power if the VSWR is too high.

If one antenna had high VSWR, say 5:1, it would introduce a small loss, a few dB, on transmit and receive. However with the protection circuit active, only the transmitter power might be reduced by a further 10 dB.

You could test the VSWR, or just add a 6 dB attenuator or a roll of cable in line with the antenna, and see if the RSSI becomes symmetrical. The suspect antenna is on the opposite side to the radio with the low RSSI reading.

A second, more subtle effect, could come simply from the output impedance match of the transmitter. While the receiver is likely quite well matched to 50 Ohms, the transmitter might have a low impedance. This wouldn't matter if the antenna match is good. However with a badly matched antenna, you could have an accidental enhancement or reduction in transmission power. This would also be an asymmetrical effect, but probably only 1 or 2 dB.

$\endgroup$
0
$\begingroup$

The difference in RSSI is easily explained by the two different antennas. The omni-directional antenna transmits the energy of the radio waves equally in all directions. (let's assume a 2 dimensional world for now. ) So think of circles emanating from a single point.

The yagi, on the other had, transmits the same TOTAL energy but loads more of that energy in the direction that it is pointing. It does this because it reflects more of it's total energy in one direction.

Hence you will receive more energy on the omni if the yagi is pointing at it. Conversely of the yagi was pointing away from the omni, the omni side would receive less.

Caveat: there is always something else going. Reflected waves can combine in positive or negative ways. Also even two products that are the same model can be different, given that each component on them is accurate within a range. etc. etc.

$\endgroup$
  • 1
    $\begingroup$ No. Antennas are reciprocal, they receive and transmit in the same way. Path reflections too. Look at the equation for path loss, it gives no special treatment to G1 and G2, they're just multiplied together, and multiplication is commutative. $\endgroup$ – tomnexus Apr 22 '16 at 9:24
  • $\begingroup$ Have to agree with tomnexus here. If the yagi was pointing away from the omni, the omni side would receive less, as would the yagi side, and by the same amount. Agree on your caveat, but we're talking a 15dB difference, which is quite significant. $\endgroup$ – epv Apr 22 '16 at 10:58

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.