Simulink: How to set parameters of AWGN Channel Block ?
13 次查看(过去 30 天)
显示 更早的评论
Hi,
I'm using Simulink to simulate an OFDM communication system with AWGN Channel, and have some problems of setting parameters.
How does Spectrum Analyzer calculate the Channel Powers ? Is it based on 1 Ohm or 50 Ohm?
In my case, I use ACPR of Spectrum Analyzer to measure the SNR of signal after channel. The SNR and Input signal power, referenced to 1 ohm (watts) of AWGN Channel Block are 20 dB and 1 watts respectively. When the power of ACPR is around 20 dB , which is equal to the SNR we set in AWGN Channel Block, the Channel Powers showed in Spectrum Analyzer is around -20 dB, which I expect to be around 30 dB for referencing to 1 watts should result in -30 dB.
Where did I do wrong? What's the relationship between Input signal power, referenced to 1 ohm (watts) of AWGN Channel Block and Channel Powers showed in Spectrum Analyzer?
![](https://www.mathworks.com/matlabcentral/answers/uploaded_files/148135/image.png)
![](https://www.mathworks.com/matlabcentral/answers/uploaded_files/148136/image.png)
Best, Ruozhu
0 个评论
采纳的回答
Greg Dionne
2015-3-20
编辑:Greg Dionne
2015-3-20
Hi Ruozhu,
You can adjust the reference impedance of the Spectrum Analyzer by going to the "Spectrum Settings" dialog, under the "trace options" panel, where you'll find the reference load.
The noise floor you have (as best as I can estimate from your plot) is -72 dBm/Hz. This is consistent with what you would expect from -20 dBW of total noise (or 10 dBm of noise) across your entire spectrum (~130 MHz). -72 + 10*log10(130e6) = 9.14 dBm = -20.86 dBW
My hunch is that your input signal at the input to the AWGN channel is actually at a power level of 20 dBm (-10dBW), not 30 dBm (or 0 dBW) as you were expecting. The fact that the drop in noise density between your original signal and the adjacent channel is 20 dB seems a coincidence.
If you remove the AWGN channel, do you still observe the same integrated power of 20dBm?
HTH
-Greg
2 个评论
Greg Dionne
2015-3-23
编辑:Greg Dionne
2015-3-23
Hi Ruozhu,
As far as I can tell, the AWGN block doesn't actually compute the input power when in the SNR mode. It just adds a noise power that is offset from the "Input signal power". If the input signal power doesn't match your actual input power, then you'll be off by whatever difference in power is there.
Maybe an exercise would help us understand what is going on.
1. Drop in a DSP Sinewave source block. Make it have an amplitude of 1, but have the same sample rate as your original signal, additionally set its output complexity to "complex".
2. In the spectrum analyzer see if you can read back a value of 30 dBm (use the "Power" spectrum type in the Spectrum Settings dialog).
3. Now put in the AWGN block. Set it for, say, 1 Watt input power and an SNR value of 20.
4. Go to the spectrum analyzer and view the spectrum. Is it what you expect?
5. Now change the amplitude of the sinusoid to 0.1. You should see the sinusoid drop by 20 dB to 10dBm, but see the same noise power as before.
6. Look at the "Input power" control. Note that it still reads 1 not 0.1^2. Change it to 0.01. Now the SNR is "correct" in the scope.
The key point is that the SNR mode doesn't actually know anything about the actual input power of the input signal. It really is just a relative power offset from the reference input power level you typed in, not the actual input power level to the block.
Hope that helps.
-Greg
更多回答(0 个)
另请参阅
类别
在 Help Center 和 File Exchange 中查找有关 Synchronization and Receiver Design 的更多信息
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!