I want to multiply two signals with AWGN=A3=ACthe result is bad if the SNR is low. I wonder if there are any methods to improve the SNR. Thank you for your help.
multiply AWGN?
Started by ●September 21, 2007
Reply by ●September 21, 20072007-09-21
Dorothy <dorothys.yuan@gmail.com> writes:> I want to multiply two signals with AWGN��the result is bad if the SNR > is low. > I wonder if there are any methods to improve the SNR. > Thank you for your help.I want to steal my neighbor's new Cadillac STS. The result is bad. How do I improve it? -- % Randy Yates % "How's life on earth? %% Fuquay-Varina, NC % ... What is it worth?" %%% 919-577-9882 % 'Mission (A World Record)', %%%% <yates@ieee.org> % *A New World Record*, ELO http://home.earthlink.net/~yatescr
Reply by ●September 21, 20072007-09-21
On Fri, 21 Sep 2007 09:49:58 -0400, Randy Yates <yates@ieee.org> wrote:>Dorothy <dorothys.yuan@gmail.com> writes: > >> I want to multiply two signals with AWGN��the result is bad if the SNR >> is low. >> I wonder if there are any methods to improve the SNR. >> Thank you for your help. > >I want to steal my neighbor's new Cadillac STS. The result is bad. >How do I improve it?Wear a mask. The shape of the mask may be important. Eric Jacobsen Minister of Algorithms Abineau Communications http://www.ericjacobsen.org
Reply by ●September 21, 20072007-09-21
Eric Jacobsen <eric.jacobsen@ieee.org> writes:> On Fri, 21 Sep 2007 09:49:58 -0400, Randy Yates <yates@ieee.org> > wrote: > >>Dorothy <dorothys.yuan@gmail.com> writes: >> >>> I want to multiply two signals with AWGN��the result is bad if the SNR >>> is low. >>> I wonder if there are any methods to improve the SNR. >>> Thank you for your help. >> >>I want to steal my neighbor's new Cadillac STS. The result is bad. >>How do I improve it? > > Wear a mask. The shape of the mask may be important.Better, don't do something stupid in the first place! -- % Randy Yates % "Maybe one day I'll feel her cold embrace, %% Fuquay-Varina, NC % and kiss her interface, %%% 919-577-9882 % til then, I'll leave her alone." %%%% <yates@ieee.org> % 'Yours Truly, 2095', *Time*, ELO http://home.earthlink.net/~yatescr
Reply by ●September 22, 20072007-09-22
On 9=D4=C222=C8=D5, =C9=CF=CE=E72=CA=B150=B7=D6, Randy Yates <ya...@ieee.or= g> wrote:> Eric Jacobsen <eric.jacob...@ieee.org> writes: > > On Fri, 21 Sep 2007 09:49:58 -0400, Randy Yates <ya...@ieee.org> > > wrote: > > >>Dorothy <dorothys.y...@gmail.com> writes: > > >>> I want to multiply two signals with AWGN the result is bad if the S=NR> >>> is low. > >>> I wonder if there are any methods to improve the SNR. > >>> Thank you for your help. > > >>I want to steal my neighbor's new Cadillac STS. The result is bad. > >>How do I improve it? > > > Wear a mask. The shape of the mask may be important. > > Better, don't do something stupid in the first place! > -- > % Randy Yates % "Maybe one day I'll feel her cold embra=ce,> %% Fuquay-Varina, NC % and kiss her interfa=ce,> %%% 919-577-9882 % til then, I'll leave her alo=ne."> %%%% <ya...@ieee.org> % 'Yours Truly, 2095', *Time*, ELO=http://home.earthlink.net/~yatescr I am not quite sure about your meaning. Should I give up the attempt?
Reply by ●September 22, 20072007-09-22
On 9=D4=C222=C8=D5, =C9=CF=CE=E71=CA=B137=B7=D6, Eric Jacobsen <eric.jacob.= .=2E@ieee.org> wrote:> On Fri, 21 Sep 2007 09:49:58 -0400, Randy Yates <ya...@ieee.org> > wrote: > > >Dorothy <dorothys.y...@gmail.com> writes: > > >> I want to multiply two signals with AWGN the result is bad if the SNR > >> is low. > >> I wonder if there are any methods to improve the SNR. > >> Thank you for your help. > > >I want to steal my neighbor's new Cadillac STS. The result is bad. > >How do I improve it? > > Wear a mask. The shape of the mask may be important. > > Eric Jacobsen > Minister of Algorithms > Abineau Communicationshttp://www.ericjacobsen.orgWhat do you mean by the "mask"? I hav never heard of it. Do you mean add a window?
Reply by ●September 22, 20072007-09-22
Dorothy <dorothys.yuan@gmail.com> writes:> On 9��22��, ����2ʱ50��, Randy Yates <ya...@ieee.org> wrote: >> Eric Jacobsen <eric.jacob...@ieee.org> writes: >> > On Fri, 21 Sep 2007 09:49:58 -0400, Randy Yates <ya...@ieee.org> >> > wrote: >> >> >>Dorothy <dorothys.y...@gmail.com> writes: >> >> >>> I want to multiply two signals with AWGN the result is bad if the SNR >> >>> is low. >> >>> I wonder if there are any methods to improve the SNR. >> >>> Thank you for your help. >> >> >>I want to steal my neighbor's new Cadillac STS. The result is bad. >> >>How do I improve it? >> >> > Wear a mask. The shape of the mask may be important. >> >> Better, don't do something stupid in the first place! >> -- >> % Randy Yates % "Maybe one day I'll feel her cold embrace, >> %% Fuquay-Varina, NC % and kiss her interface, >> %%% 919-577-9882 % til then, I'll leave her alone." >> %%%% <ya...@ieee.org> % 'Yours Truly, 2095', *Time*, ELO http://home.earthlink.net/~yatescr > > I am not quite sure about your meaning. > Should I give up the attempt?Dorothy, In general, purposely introducing noise into a processing step, especially as a multiplication, is not a good idea [1]. You've given us four sentences with very little context or orientation as to what it is you're trying to accomplish. Why don't you back up and present us with the problem you're trying to solve instead of giving us the tail end of an implementation of a solution that didn't work? --Randy [1] Dithering a signal prior to quantization is an exception, although this does not involve multiplication. -- % Randy Yates % "Remember the good old 1980's, when %% Fuquay-Varina, NC % things were so uncomplicated?" %%% 919-577-9882 % 'Ticket To The Moon' %%%% <yates@ieee.org> % *Time*, Electric Light Orchestra http://www.digitalsignallabs.com
Reply by ●September 22, 20072007-09-22
On Fri, 21 Sep 2007 07:57:39 +0000, Dorothy wrote:> I want to multiply two signals with AWGN，the result is bad if the SNR > is low. > I wonder if there are any methods to improve the SNR. > Thank you for your help.You need to carefully define what your signals of interest are, and what a 'good' result means. Depending on how you treat your signals there may be ways to improve the SNR, or not. The first and best, but perhaps hardest way, is to figure out what you're trying to estimate, then find the best fit to that estimate. Presumably you've got two transmitted signals, A and B, and you're trying to estimate (A)(B), but what you're receiving is A + AWGN and B + AWGN -- is that correct? Your second-best approach would be to find the best estimates of A and B, and multiply those estimates. If A and B are bandlimited then you should be able to use Wiener filters to estimate them. If that raises the SNR enough, then you don't have to go invent a bunch of new science to do the estimation directly. This is probably what I'd try first, at least on paper, to see if it would get me what I needed. -- Tim Wescott Control systems and communications consulting http://www.wescottdesign.com Need to learn how to apply control theory in your embedded system? "Applied Control Theory for Embedded Systems" by Tim Wescott Elsevier/Newnes, http://www.wescottdesign.com/actfes/actfes.html
Reply by ●September 23, 20072007-09-23
Thank all of you for your patience. I am not a native speaker and maybe cannot express myself on a professional question exactly. Still I want to try. I've got two different signals x(t) and y(t) x(t) = exp(j*2*pi*(f1*t+f2*t)) ; y(t) = exp(j*2*pi*(-f1*t+f2*t)) ; The signal-to-noise ratios of x(t) and y(t) are similar. I want to derive f2 from x(t) and y(t), if there is no noise , I can derive f2 simply from x(t)*y(t) but if the SNR is low,the result of x(t)*y(t) is unreliable. Now I wonder if I should find some ways to improve the SNR or cast away the method completely. Dorothy
Reply by ●September 23, 20072007-09-23
Are you filtering the product? - decide on a sufficiently wide frequency window, so that it must contain the product - filter out all energy below / above the window This might give enough SNR improvement. FFT and peak search would also do the job, but might be considered a "brute force" solution. Or use a PLL on the product, see Proakis "Carrier and Symbol Synchronization". -mn
