Extra point in constellation point in OOFDM

[Gokul Boro .]

Hi everybody,
I am confused with this design (attached). There is on extra constellation point. How can I adjust the parameters so that I can get rid of this point. Also I tried the same for 4QAM and the constellation seems to be ideal even though I have used fiber for transmission. Can anyone help me to know whether it is possible to get such constellation in a non-ideal system?

Attachments:

1. BICR-OOFDM-System-with-OFDM-block-set.osd

[alistu]

Hi Gokul,

I can see a corrupted signal at the receiver and not a constellation with just one extra point. All the signals have been gathered at the center. I am attaching a snapshot of the constellation I can see here. Can you please check to see if the design you have attached is actually what you intended to attach?

Regards

Attachments:

1. 

   corrupt.jpg
   

[Dr. Dhiman Kakati]

Dear Gokul I also dont understand about this extra point, What may be the reason for this.

Regards

[Gokul Boro .]

I have downloaded the file I have downloaded and simulated it. The design seems to be the one which I intended to attach. The snapshot of constellation is attached along with the design.

Attachments:

1. 

   Capture1.png
   

1. BICR-OOFDM-System-with-OFDM-block-set1.osd

[alistu]

Thanks for the snapshot and reattaching the file. I ran the simulation again and there was no difference in the constellation. Dhiman also went through the same thing. It is strange that although you have used a sequence of length 16000, not even one point has fallen out of the exact point it has to be in (except for the one in the center of course). Which version are you using Gokul?

[Gokul Boro .]

Hi Alistu,
I am using Version-13.0.3. Did u get the constellation as the image I have attached herein? And what does that point at the center mean?

[alistu]

No, I got something more or less the same as before. While I cannot say anything for certain, I suggest that you increase the bandwidth of the filters used in the design, as sometimes the small cut-off frequency causes some symbols to be completely filtered and therefore show themselves as a zero point in the constellation diagram. Again I am not sure if this helps.

[Damian Marek]

Hey Gokul,

I saw a couple errors in your design that I fixed and removed portions because I did not understand how they were being used! But you should be able to add them back if done correctly (and its a working design). You should use the BER Test Set for all things OFDM because it takes into account the pilot and training symbols. Also your quadrature demodulator had a low pass filter of 10 MHz which is much too low for the ~2 GHz bandwidth signal.

Carefully read the technical background of the OFDM Modulation and Demodulation and how it pertains to the BER Test Set because most of your errors are explained in the Help file already.

Regards

Attachments:

1. 

   Capture2.png
   

1. BICR-OOFDM-System-with-OFDM-block-set-DMEdit.osd

[Gokul Boro .]

Thank you very much Damien for correcting the errors.

[Gokul Boro .]

Hi Damien, I tried simulating the .osd file that you modified. But the constellation seems to be ideal as was before except for the additional point. The image is attached. I have modified nothing in the .osd file you have attached. If you got the constellation as that of the image you have attached attach from the system you have attach then please inform here. Alistu got a different constellation from the system I attached and now I am getting a different one from your design (constellation if the parameters were not modified for getting above image of constellation.)
thank you.

Attachments:

1. 

   IDEAL.png
   

[alistu]

Actually I was wrong when i said the constellation should not be so ideal. I hadn’t noticed you were using OptiSystem 13 OFDM components. The signal in OFDM demodulator goes through a process of decision making and falls in one of the main points it should be in. To see the constellation before this process, you can go to project browser and choose OFDM demodulation component and see it in the results.

[Damian Marek]

Alistu is right. Project Browser->OFDM Demodulation->Graphs->Constellation diagram before carrier estimation

[Gokul Boro .]

The constellation visualizer output that i have used in the system shows the same result even if I change the parameters. But if I replace the visualizer with another one then the visualizer shows different visual. Why does it happen? Also Can I use the BER test Set with OS12 block?

[alistu]

I don’t know why changing constellation visualizer makes a change to the signal, but as for what you have asked about using BER test set with OS12 OFDM, I believe you can do that by putting parameters like training symbols and pilot signals to zero and matching other parameters to those of OFDM component.

[Gokul Boro .]

To some extent I learned to use OS12 block. But the new block in Version 13 seems to be much more complex. If we could have used PRBS generator instead of BER Test Set then it would have been simpler. Now it will take more time to learn using this block. On the other hand there is no tutorials for this Software.

[alistu]

I agree that is is a bit complicated and I also sometimes find problems working with it, but the OFDM samples available in OptiSystem samples help you realize how other parameters (for example in BER Test Set) need to be set according to the OFDM parameters. As for the BER Test Set, do you have any question in particular?

[Gokul Boro .]

Thank you Alistu,
Is there any sample design which uses OFDM in version 13 in the web which works properly? If so please provide the link.
Also I would like to know if these designs are possible in Optisystem-
1.DMT (Discrete Multi Tone), where half of the inputs of the IFFT are the complex conjugate of the other half so a real OFDM signal is created at the transmitter output
2. Zero Padding, to create a gap between the OFDM signal and the DC component.

[alistu]

You’re welcome. If by “any sample design which uses OFDM in version 13” you mean any file other than OptiSystem samples files (that work properly), unfortunately I don’t know any.

About the other question, I am not really sure, but you can make them using Matlab. I don’t think the first one is available. How is the second one you have mentioned different from the current position array capability which is available in OS 12 and OS 13 OFDM?

[Author]

Posts