April 12, 2015 at 3:57 pm #19601
I`am trying to simulate LTE signal OFDM 64QAM.
Damian or somebody, please, could you help me with my simulation. I have a problem with:
1) Electrical Constellation Visualizer shows me strange values of EVM (Error Vector Magnitude)/Q Factor
2) I haven
t seen any results on BER Analyzer and I have been trying to use BER Test Set, but it hasnt worked too.
version of OptiSystem 18.104.22.168
P.S. I`m sorry, but how can I upload my simulation? Capacity is about 370 MB.
That`s why I put this link: https://www.idrive.com/idrive/sh/sh/f8h8r1v1k9
April 13, 2015 at 10:50 am #19645
Generally the project files are much smaller, I think you set the Option to Save Monitor Data in the Tools->Options window. By unchecking that your project file will be much smaller.+2
April 13, 2015 at 12:28 pm #19648
The New file of simulation is attached.
April 13, 2015 at 2:30 pm #19661
The issue you have with the BER Analyzer is more of a misunderstanding. The BER Analyzer approximates the noise from the detected noise of the output signal and the characteristics of the eye diagram. In OFDM the noisy OFDM symbols are detected (perhaps not correctly) converted to an NRZ format, but now there is no noise. The more appropriate way to determine the BER is to use the BER Test Set that directly compares the input and transmitted sequence.
I’m not sure what is happening to the constellation diagram calculating the EVM’s, I’ll take a closer look.
April 13, 2015 at 2:52 pm #19662
The reason for the constellation diagram errors comes from the loss that the signal experiences. The constellation diagram needs the correct positions and sizes of the symbols. You can either introduce a gain component so the electrical signal reach the same amplitude as the input signal or you can scale the targets in the component properties of the constellation visualizer. Change the Distance scale, I target scale and Q target scale to 0.00485714 for a more realistic answer.
April 16, 2015 at 5:24 pm #19842
Thank you for the help!
I did everything as you said, but I still can’t receive the normal results. I’ve tried different options, but nothing happens. I hope for your help. Very grateful to you in advance!
1) I Connected Ber Set Test, but it shows a huge BER value. I changed many different values of a signal, and the BER value is always or 0,5 or 1.
2) I made everything as you told, but the value of EVM is still bad. Please explain in more detail what it means “You can introduce a gain component so the electrical signal reach the same amplitude as the input signal”. You mean, that the signal output amplitude from QAM modulator has to be the same, as output signal on QAM demodulator?
3) I would like to test the dependence Constellation diagram on the SNR, so after QAM modulator I added Electrical atenuator, but it does not affect the Constellation diagram. In what there can be a problem? I noticed, that with Electrical Power Meter vusualizer, there is in general no noise, perhaps the problem in it?
April 17, 2015 at 9:02 am #19860
Do you have access to the new OFDM components released with OptiSystem 13, I find them much easier to work with and troubleshoot. If you do I’ll just go ahead and replace them.
April 17, 2015 at 10:02 am #19867
Ok, thank you. You can do it in OptiSystem 13. I will download a trial version.
April 22, 2015 at 4:07 pm #20059
Alright found out what the problem was here! The quadrature mod/demod pair upconverted the signal to 1800 MHz which is MUCH larger than the bit rate and the subsequent sample rate, or atleast the one I was using in my own file. Using a high enough sample rate so that it is greater than than 1800 MHz worked for me. All that is missing is the optical link in the middle.
April 24, 2015 at 10:17 am #20103
Thank you very much, I continue to work on my project, but all attempts had practically no results. Could you please tell me, channel bandwidth can be configured through the bit rate? I chose 60 mbit/s (10 MHz Channel bandwidth) due to the best constellation diagram result, but is slightly preferable to have Channel bandwidth 20 MHz for LTE signal.
April 27, 2015 at 11:29 am #20170
Hello! Finally , I am ready with optical part. EVM values is more realistic now, but BER values are still very high. Please, help me, I have some questions for you. I hope you can help me.
1) In optisystem 12 the noise can be seen and constellation diagram isn’t ideal ( see attached image). In the 13 version of Optisystem , the resultant constellation diagram is excellent, why? Where the noise is? EVM is 8-10%, but in the diagram this fact isn’t shown up.
2) Why in the electrical part Electrical Attenuator has no effect on EVM, even at 100 dB of noise, but in optical part
Optical attenuator has utterly faint effect on EVM.
3)How can I calculate SNR? I want to draw a graph SNR/ EVM
April 29, 2015 at 6:01 pm #20224
Some of the problems are overcome, but some issues remained unresolved.
1) It is necessary to reduce the value of BER and EVM
2) The value of EVM on the OFDM modulator does not correspond to the constellation diagram, why the diagram looks completely perfect? In optisystem 12 it was entirely different. (see attached image).
May 6, 2015 at 8:54 am #20483
You will notice a difference between the older OFDM components and the newer version because the constellation diagram is provided by the OFDM Demodulator itself. The output of the OFDM Demodulator is now an Mary sequence, so of course using this signal to regenerate a constellation will appear to have no noise, since the signal has been detected.
If you use the project browser at the bottom left of the window and navigate to the OFDM Demodulation component. Take a look at the different figures in the graphs folder.
June 25, 2015 at 12:11 pm #21915ferasParticipant
hi Nikolay Volkov ,
can you please attach the simulation with optisystem 12 .
i am working on a system like that but with signal equal to 2.6 ghz ,
and thank you
May 2, 2015 at 2:58 am #20275Ajay VyasParticipant
There are some error to download from the link. Please upload again
May 2, 2015 at 9:18 am #20281
May 6, 2015 at 8:09 am #20481
Could anybody help me?
- You must be logged in to reply to this topic.