- This topic has 16 replies, 6 voices, and was last updated 9 years, 6 months ago by Abhishek Shrama.
-
AuthorPosts
-
-
May 25, 2015 at 2:16 am #21120Abhishek ShramaParticipant
@Damian
I am facing a problem with my system. Whenever I make a duplicate copy of system and simulate it, the results on BER analyzer gets changed. even I have unchecked the random seed generation still the problem persists and I noticed a changes from 0.1 to even 5 in quality factor.
I am not able to make graphs due to this problem. Even I checked in some of examples in Samples the problem lies there also.
What is the reason that when I duplicate a layout or even select all and paste it onto a new layout the results change.
What to do to overcome this problem and make graphs accurately.
I am attaching the file in attachment. Please do reply me fast. -
May 25, 2015 at 9:56 am #21136Damian MarekParticipant
You correctly disabled the generate random seed for the bit sequence generator, but there are other sources of random noise in your system. For example, the Laser and Photodetector PIN.
Regards
-
May 26, 2015 at 12:18 am #21150Abhishek ShramaParticipant
Thankyou Damian.
It solved my problem. but is it authentic to disable all noises(Signal- ASE, ASE-ASE, Thermal and shot noise) ?
and what is the I have one more doubt regarding it. The noise generation is random. Then why does it not change everytime I simulate the system and changes only when I change the layout?
and What should I prefer? whether drawing graphs with random readings or disabling all the noises and the take the readings?-
May 26, 2015 at 2:12 am #21155Alessandro FestaParticipant
Have you tried to increase the sequence length?
-
May 26, 2015 at 8:43 am #21168Damian MarekParticipant
My simulation changes with each calculation if I have the generate new seed box checked. It might be possible that you are using an older version of OptiSystem?
Regards
-
May 26, 2015 at 11:04 am #21175Alessandro FestaParticipant
Hi Damian, if you increase the sequence length to 512 the change between iterations becomes very small…is this correct?
-
May 26, 2015 at 11:59 am #21176Damian MarekParticipant
The variations will become smaller, but how small depends on the actual BER of the system. A noisier system will be more susceptible to random changes in the noise and there will be larger fluctuations in the calculated BER.
-
May 27, 2015 at 2:17 am #21189Alessandro FestaParticipant
Ok thanks Damian!
-
May 27, 2015 at 5:22 am #21192Abhishek ShramaParticipant
Thankyou damian. You are always so very helpful.
It solved the problem. I was missing out the random seed boxes checked in PIN and laser. Thats why I was getting results changed everytime. Now I got over it and getting constant results.
-
-
-
June 4, 2015 at 3:22 am #21373GURINDER SINGHParticipant
I am also getting same problem…. random seed index have to be made zero??????
-
June 4, 2015 at 3:45 am #21374Abhishek ShramaParticipant
yes GURINDER you have to uncheck the random seed generation not only from PRBS but also from all components like PIN LASER etc. Then you will get constant results to make decisions.
-
June 4, 2015 at 7:47 am #21386GURINDER SINGHParticipant
okk….i have unchecked it….is the random seed index also made to be zero in each case??
-
June 4, 2015 at 10:43 am #21399K. Esakki MuthuParticipant
Thank you Guys, I too got the same problem and now I got cleared my doubt.
Thanks to the optisystem forum
-
June 4, 2015 at 1:39 pm #21401GURINDER SINGHParticipant
have u made random seed index also zero?????
-
June 5, 2015 at 12:10 am #21402Abhishek ShramaParticipant
Hello Gurinder,You can make the value of random seed generator zero.If you will have same problem ater setting seed to zero then do same as Damian wrote in previous reply.
-
June 7, 2015 at 9:40 pm #21445Dr. MohammadMahdi AriannejadParticipant
same problem goes here!!!!!
-
June 8, 2015 at 1:13 am #21458Abhishek ShramaParticipant
@ MohammadMahdi Ariannejad
Please read the discussion above and follow the steps of switching off the random seed generation from generators of random noise.
-
-
AuthorPosts
- You must be logged in to reply to this topic.