In the context of wireless communication, I would like to know the relevance of power at the receiver side. In literature it is mentioned that received power should be maximum. This seems fine intuitively but I don't know where exactly this matters.

Does the power at the receiver matter at the RF or the baseband stage?

If two signals at the receiver (e.g. multipath) add constructively then would there be any issue with the baseband data in terms of position in the constellation space?

A: The receiver adds noise to the receive signal - so more signal power means more signal-to-noise-ratio and that is good for both. However: Be careful not to have too much power in the receiver, or you will boost intermodulation products (caused by non-linearity of the receiver).

Q:If two signals at the receiver (e.g. multipath) add constructively then would there be any issue with the baseband data in terms of position in the constellation space?

A: When two sine-waves with the same frequency, but different phase add up, you get another sine of the same frequency, but different amplitude and phase. if your modulation signal is narrow-band (that means the modulation bandwith is small compared to the carrier frequency), the worst you will get is a constant phase shift (if the multipath situation is not changing very quickly). If it is wide-band you might run into other problems which I'd have to think about for a little longer :)

Q: Does the power at the receiver matter at the RF or the baseband stage?A: The receiver adds noise to the receive signal - so more signal power means more signal-to-noise-ratio and that is good for both. However: Be careful not to have too much power in the receiver, or you will boost intermodulation products (caused by non-linearity of the receiver).

Thanks. That makes sense about SNR but question still is, beyond a point how higher SNR is going to make any difference. I would assume that there is a point (either a specific design parameter or in terms of statistics) beyond which increasing the signal power does not serve any purpose.Also, on your comment about being "good for both", I assume you mean both RF & baseband (and Fred Marshall clarifies additionally Antenna). What does additional power help once RF stage is done processing? In other words, if received power is up, does it make any difference to the IQ values that go to the baseband?

Q:If two signals at the receiver (e.g. multipath) add constructively then would there be any issue with the baseband data in terms of position in the constellation space?A: When two sine-waves with the same frequency, but different phase add up, you get another sine of the same frequency, but different amplitude and phase. if your modulation signal is narrow-band (that means the modulation bandwith is small compared to the carrier frequency), the worst you will get is a constant phase shift (if the multipath situation is not changing very quickly). If it is wide-band you might run into other problems which I'd have to think about for a little longer :)

I am talking about two signals adding constructively in which case signal power increases. If such multiple constructive additions happen, I guess it increases signal power but does it cause any issue due to too much received power

At the RF front end of the receiver, there is a Low Noise Amplifier (LNA).

Sometimes, there is both an analog and a digital AGC that adjusts the signal to bring into the right range for processing.  It is unlikely that you will cause any overload, unless the TX and RX are very close together.

So most of you concerns related to signal level are usually taken care of by other analog or digital processing.

David

to explain my question further, please refer to the figure below. Both the signals are same except for the amplitude/power at the receiver. In this case, what difference does it make to the receiver and at what stage?

Also, does it have any impact on the baseband side, which I believe is much critical in order to get proper data

My immediate response it in relation to the baseband signal to which you refer.

The level of the analog signal will also define the signal to quantization ratio.  When the signal is sampled to get it down to baseband, my assumption is that two A/Ds will capture the I and Q components.  If the signal is rather small, then the quantization noise of the sampling process will add to all the other channel noise and further reduce the overall SNR.  The quantization noise will have very different statistics that will not match the AWGN with which the receiver is usually optimized to deal.

So, my overall thinking would be that the signal needs to be as large as possible to fit into the bit resolution of the A/D converter without ever clipping or experiencing any system non-linearity.  This will guarantee the best SNR with minimum quantization noise.

If the capture of the A/Ds is as good as possible, then everything else can be adjusted in the baseband software/processing.

The one other thing that comes to mind is knowing what all the gains in the system are.  If you need to know what signal level you actually have (e.g. for carrier detect) when you get down to baseband, then you will need to keep track of all the gains and attenuations through which the signal has passed so that you can account for them in the baseband processing.  As I mentioned before, there may be an analog AGC in the pipeline.  This is very often controlled by a feedback process from the digital baseband signal level analysis.

My 2 cents.

David

Dear dgshaw6,

The waveforms shown are the RF or antenna input. You have assumed that these signals are what they look like when baseband data is transformed into IQ data.

In fact, my additional question is related to the baseband side. If I have 2 signals of same shape but different amplitude at RF input then how do they look like when the data is converted to IQ. My intention of this question is to understand if the SNR is adequate enough then does signal power make any difference the way data is seen at the baseband if the power levels of 2 signals are different but waveforms are same.

The only difference for the baseband will be the signal to quantization ratio SQR.

For most modulation schemes though, there is an optimal received signal level.

For simpler schemes like FSK or CPM the issues with quantization are smaller.

For complicated modulations like OFDM, then the peak to average ratio becomes of significant concern.  We never want to clip any peaks no matter what modulation technique is in play. (Some might disagree about the damage that would be done to FSK, but no clipping is a general rule of thumb.)

For all modulations, the receive signal processing will have an ideal level at which the performance will be optimal.  As long as the combined SNR and SQR meet receiver requirements at the A/D then everything else is done in the baseband signal processing to bring the signal level to the ideal for the best algorithm performance.

Ignoring noise, on the baseband side the signals will have the same shape but different amplitudes.  Paying attention to the noise, the signals will have different amplitudes, but the impressed noise amplitude will be the same -- so the SNR will be lower for the smaller signal.

Typically, the signal strength of received signals in an RF link will vary widely, and sometimes rapidly.  It's the job of the system designer to design signals that make it easy to deal with this variation, and for the receiver designer to make the receiver robust in the face of this variation.

I well-designed receiver should be pretty robust in the face of the intended receive signal's power being very strong.  If the wrong combination of interfereing signals is strong in relation to the signal strength then that's a different story -- hunt down references to receiver dynamic range for more details.

dudelsound is exactly right about the issues of SNR.

Everything about the receiver should be designed to maximize the SNR and in the "N" he has correctly added the issues with non-linearity.

The first issue that would come to mind for wider band signals is that some spectral components may add constructively, while other parts of the spectrum may add destructively.  There are many techniques used to solve these issues in the signal processing of the baseband signal.

In the wide-band case, if the bandwidth is wide because the symbol length is short, then a multipath delay that's a significant proportion (or even longer than) a symbol time will cause intersymbol interference.

If the bandwidth is wide because the system is direct-sequence spread spectrum, then there will be two correlation peaks at the receiver, which the receiver must either choose between, or be built to understand how to deal with (i.e., a rake receiver).

Does the power at the receiver matter at the RF or the baseband stage?

Yes, it does.  At the most fundamental level, if there's no signal then there's nothing to detect, process, communicate.  If there's lots of signal then there is and the more the better.  That's really all there is to it.  

At which stage?  Well usually the issue is at the antenna.  So maybe that means RF to you.  Or, in the receiver design there is usually an RF section first that is followed by other stages - which it appears you refer to.  It's important for low-signal systems (e.g. space communications) for the first stage to be low noise.  Beyond that it's more controllable and subject to good design.

If two signals at the receiver (e.g. multipath) add constructively then would there be any issue with the baseband data in terms of position in the constellation space?

If they add constructively then that means they are perfectly (or nearly so) in phase. Some communication systems may rely on this.  The most common way is through the use of multielement antennas.  A dipole is the simplest kind and acts like 2 elements.  So, when the element spacing is 1 wavelength, they add and when it's 1/2 wavelength, they subtract.

If the delay is limited to a wavelength (which is determined by the speed of light and the frequency), then the amount of delay between "arrivals" is trivial.  Possibly, for very high data rates, one has to be concerned with this sort of thing.  In that case I can but imagine that the system design deals with it.    So, I'd say, in general: no.

Dear Fred,

Thanks a lot. My follow up question is above, if you could kindly answer ...

Both the signals are same except for the amplitude/power at the receiver. In this case, what difference does it make to the receiver and at what stage?

You have only shown the "signals" and it's unclear whether these include the noise.  I'll assume, in view of the question and for purposes of discussion, that this is the total signal at the antenna. So, you have to add the noise of the receiver amplifier stages (usually the first stage is the most important for this) and the bandwidth of the amplifier/filtering perhaps.  Usually the noise is specified in terms of "noise spectral density" which is per Hz or sqrt Hz.  So, the wider the bandwidth, the higher the noise level.
So assuming the noise is unchangeable then the smaller signal will be compared to the added noise and the SNR will be lower than for the larger signal.

Also, does it have any impact on the baseband side, which I believe is much critical in order to get proper data

Many things would "have impact on the baseband side".  So, one simple view would be to simply say "no".  
But if there is any bandpass filtering involved then the SNR could be improved by removing noise outside the signal spectrum - however that may be done.  

And, I suppose it might be worthwhile to say that it's likely easier to do such filtering at baseband because the relative bandwidth of the signal is larger - requiring less sharp filters or lower "Q" for the same effect.  It all depends on the system design.  This is not to say that there aren't better or best designs overall that do the filtering elsewhere in the processing chain.  I don't know but it certainly could be.  There are lots of clever approaches out there.