I need to determine minimum working SNR for transceiver system over a fixed distance. I change the Tx power till the Rx cannot make sense of the message. The problem is to estimate the SNR at the receiver. I've two main questions 1. Is it possible to estimate the SNR reliably from the spectrum of the received signal? 2. Should the SNR be computed before matched filtering or after?
SNR estimation before or after matched filtering
Started by ●October 30, 2011
Reply by ●October 30, 20112011-10-30
>I need to determine minimum working SNR for transceiver system over afixed distance. I change the Tx power till the Rx cannot make sense of the message. The problem is to estimate the SNR at the receiver. I've two main questions>1. Is it possible to estimate the SNR reliably from the spectrum of thereceived signal?>2. Should the SNR be computed before matched filtering or after? >It is more practical to inject a tone at Tx and measure SNR at Rx since then you can define the bins for signal and noise separately. As such matched filter is irrelevant. If you don't use tone then you will find it difficult to separate noise from signal. If you do matched filtering and decimate down then that makes it even more difficult to discern signal from noise and I would rather depend on BER at that level. Regards Kadhiem
Reply by ●October 30, 20112011-10-30
On Oct 30, 6:53�am, mobi <mob...@gmail.com> wrote:> I need to determine minimum working SNR for transceiver system over a fixed distance. I change the Tx power till the Rx cannot make sense of the message. The problem is to estimate the SNR at the receiver. I've two main questions > 1. Is it possible to estimate the SNR reliably from the spectrum of the received signal? > 2. Should the SNR be computed before matched filtering or after?Is your system spread spectrum? All bets are off if it is. But for simple systems I think the answer to both questions is yes. If your matched filter is a typical IF filter in hardware, then I think yes to #2. If the matched filter is in software, then the problem will be more difficult. I don't think there is a one answer fits all architectures.
Reply by ●October 30, 20112011-10-30
On Sun, 30 Oct 2011 03:53:04 -0700 (PDT), mobi <mobien@gmail.com> wrote:>I need to determine minimum working SNR for transceiver system over a fixed distance. I change the Tx power till the Rx cannot make sense of the message. The problem is to estimate the SNR at the receiver. I've two main questions >1. Is it possible to estimate the SNR reliably from the spectrum of the received signal?Yes, but remember that the PSD that you measure within the signal bandwidth is S+N, so that the height of a signal above the noise floor is (S+N)/N not S/N.>2. Should the SNR be computed before matched filtering or after?It can be either, you just need to be careful of what you're doing. e.g., the noise BW is different after the matched filter than before, but SNR can be accurately estimated either way. Eric Jacobsen Anchor Hill Communications www.anchorhill.com
Reply by ●October 30, 20112011-10-30
mobi wrote:> I need to determine minimum working SNR for transceiver system over a > fixed distance. I change the Tx power till the Rx cannot make sense > of the message. The problem is to estimate the SNR at the receiver. > I've two main questions 1. Is it possible to estimate the SNR > reliably from the spectrum of the received signal? 2. Should the SNR > be computed before matched filtering or after?Neither. To get an accurate estimate of SNR, use a noise generator and compare noise power to transmit power. Vladimir Vassilevsky DSP and Mixed Signal Design Consultant http://www.abvolt.com
Reply by ●October 30, 20112011-10-30
> > >mobi wrote: > >> I need to determine minimum working SNR for transceiver system over a >> fixed distance. I change the Tx power till the Rx cannot make sense >> of the message. The problem is to estimate the SNR at the receiver. >> I've two main questions 1. Is it possible to estimate the SNR >> reliably from the spectrum of the received signal? 2. Should the SNR >> be computed before matched filtering or after? > >Neither. > >To get an accurate estimate of SNR, use a noise generator and compare >noise power to transmit power. > > >Vladimir Vassilevsky >DSP and Mixed Signal Design Consultant >http://www.abvolt.com >But why add noise when you want to measure your Tx/Rx channel noise? Kadhiem
Reply by ●October 30, 20112011-10-30
On 10/30/2011 12:59 PM, Vladimir Tnediputs wrote:> > Neither. > > To get an accurate estimate of SNR, use a noise generator and compare > noise power to transmit power. > >Vlad Tnediputs, I don't know what you were thinking when you wrote this.... I guess it must be tongue-in-cheek, eh? 1) Set up noise generator with known power output. 2) Set up signal generator with known power output. 3) Take the power ratio of S to N from the set, known quantities. Was that it???? It doesn't help the OP that much though.... But, you know that. |-)
Reply by ●October 30, 20112011-10-30
On 10/30/2011 8:57 AM, Eric Jacobsen wrote:> On Sun, 30 Oct 2011 03:53:04 -0700 (PDT), mobi<mobien@gmail.com> > wrote: > >> I need to determine minimum working SNR for transceiver system over a fixed distance. I change the Tx power till the Rx cannot make sense of the message. The problem is to estimate the SNR at the receiver. I've two main questions >> 1. Is it possible to estimate the SNR reliably from the spectrum of the received signal? > > Yes, but remember that the PSD that you measure within the signal > bandwidth is S+N, so that the height of a signal above the noise floor > is (S+N)/N not S/N. > >> 2. Should the SNR be computed before matched filtering or after? > > It can be either, you just need to be careful of what you're doing. > e.g., the noise BW is different after the matched filter than before, > but SNR can be accurately estimated either way. > > > Eric Jacobsen > Anchor Hill Communications > www.anchorhill.comYep. It's not something to just jump into a quick solution. Try this "thought experiment": Assume that the receiver prefiltered output noise is broadband Gaussian, therefore white in the passband of the receiver. Often this is a good model anyway. And, if this assumption isn't a good one, you can figure that out later. Measure the prefiltered receiver output noise and assume the spectral character above. [White noise power into the filter] Filter the noise. Measure the filter output noise power. The result should be readily estimated as well as measured. If the results don't agree then there's something wrong with the analysis or the noise wasn't white. Now inject a signal at the receiver output. Because you know the actual noise power at the receiver output, you can create an artificial / test case of signal this way. Now you can change the signal level to get the threshold you're looking for and will be able to state the SNR at the receiver output prefiltered. Obviously, the filtered output SNR should be higher than the prefiltered SNR - by a known amount if the noise is white. If the noise isn't white then you can calculate the result for pink noise or ... whatever. The question this raises is why do you want to know the receiver output SNR at threshold? Presumably you already know the noise level, the transmission loss and want to know the signal power necessary to "just work" and will add some margin. Is that it? Fred
Reply by ●October 30, 20112011-10-30
Fred Marshall wrote:> On 10/30/2011 12:59 PM, Vladimir Tnediputs wrote: > >> >> Neither. >> >> To get an accurate estimate of SNR, use a noise generator and compare >> noise power to transmit power. >> > Vlad Tnediputs,???> I don't know what you were thinking when you wrote this.... I guess it > must be tongue-in-cheek, eh?This is how professionals measure S/N threshold for a communication system. As opposed to stupidents, professors, lamers and amateurs.> 1) Set up noise generator with known power output. > 2) Set up signal generator with known power output. > 3) Take the power ratio of S to N from the set, known quantities. > > Was that it???? > > It doesn't help the OP that much though.... But, you know that.OP = idiot. The only help for him could be getting born in a new incarnation.
Reply by ●October 30, 20112011-10-30
> > >Fred Marshall wrote: > >> On 10/30/2011 12:59 PM, Vladimir Tnediputs wrote: >> >>> >>> Neither. >>> >>> To get an accurate estimate of SNR, use a noise generator and compare >>> noise power to transmit power. >>> >> Vlad Tnediputs, > >??? > >> I don't know what you were thinking when you wrote this.... I guess it>> must be tongue-in-cheek, eh? > >This is how professionals measure S/N threshold for a communication >system. As opposed to stupidents, professors, lamers and amateurs. > >> 1) Set up noise generator with known power output. >> 2) Set up signal generator with known power output. >> 3) Take the power ratio of S to N from the set, known quantities. >> >> Was that it???? >> >> It doesn't help the OP that much though.... But, you know that. > >OP = idiot. The only help for him could be getting born in a new >incarnation. > >It looks like that there are two issues raised here so far in this post; either measurement of SNR of received signal or it is about noise threshold of the comms system. I think the OP was about first. kadhiem
