how to find Instantaneous SNR

[Hassan313]

As we know that

Average SNR =(mean square value)*(instantaeous SNR)

is it possible to find out instantaneous SNR if mean square value and Average SNR are given?? plz guide me as I m really stuck

 

[vedagarbhamsandeep]

As we know that

Average SNR =(mean square value)*(instantaeous SNR)

is it possible to find out instantaneous SNR if mean square value and Average SNR are given?? plz guide me as I m really stuck

If mean square value and Average SNR are given,

(instantaeous SNR) = Average SNR / (mean square value) .. LOL !

 

[Mityan]

Very strange formula. Can you tell where did you see it?

 

[Hassan313]

If mean square value and Average SNR are given,

(instantaeous SNR) = Average SNR / (mean square value) .. LOL !

here the mean square value of the channel fading coefficient is taken e,g; if 'x' is the fading coeffecient then mean squre value of h can be written as
E(h^2),where E is for Expectation,Eb is Signal Power,No is Noise Power.

(Eb/N0)_avg= E(h^2) * Eb/No_instantaneous
I hav seen this formula in proakis 's book "digital Communication" & Salim Alouini's Book "digital Communication for Fading channels".
Now if I'm given a threshold for BER in a specific system then there must be SNR corresponding to that BER,i need to find that SNR,but most of the formulas for probabilty of error for fading channels given in literature use average SNR ,then how can i get Instantaneous SNR for that???

- - - Updated - - -

sorry x is written mistakenly the channel fading coffiecient is 'h'

 

[ChakerQ]

Hi,

Instantaneous power of the signal x is x².
Average power of the signal x with N samples is (1/N)*sum(x²).

take a look here
**broken link removed**

Regards.

Chaker

 

[Hassan313]

thanks Alot

 

[D.A.(Tony)Stewart]

Since random noise is in-determinant, you can not calculate the error for 1 bit. You can only say the probability of error.

In the 80's we could predict Soft BER and Hard BER on magnetic HDD's using 1 full track of 10KB ( don't laugh) and with delay-lines reduce the window of data recovery top measure margin and measure BER of 1e-5 and predict BER of 1e-10 for soft errors and 1e-12 for hard errors.

This usually involved plotting the margin to error at 1e-5. 1e-6, 1e-7 and extrapolating the margin to zero to see what the error rate was for 1e-10 and 1e-12. The slope of the curve translates into SNR and the asymptote near 1e-2 is a result of clock jitter, and quantization error (ISI, slicer offset, {group delay distortion in filter, lack of precompensation, }asymmetry in rise time, etc,etc), since this error is predictable < 100 bits.

Worst case pattern for MFM channel recording was 011011011011 for ISI. or 6DB 6DB...
A different pattern was worst case for RLL encoding.

The results were tested over the entire medium for channel location noise, adjacent track interference, media defect noise and compiled into a BER margin budget for establishing reliability of data communication and mapping defect locations, which were pattern dependant.

Special testers include the Cambrian (boat anchor written in Forth) and Dr Guzik's PMA and many others.