Home Forums GENERAL what is the secret ?

what is the secret ?

Profile Photo
(joined April 2014)
Participant

really what is the secret in this program , i designed tow same simulation and i got tow different result < what is the reason?

Responses (10):

    • #21429
      Profile Photo
      feras
      Participant

      really what is the secret in this program , i designed tow same simulation and i got tow different result < what is the reason?

    • #21435
      Profile Photo
      Ravil
      Participant

      Hello Feras,

      If I understood you situation correctly, you designed two the same schemes on two different layouts. If, according to your question, you have two different sets of results, then the main reason can be the sequence generated and transmitted by PRBS in your schemes. You can check this by disabling random sequences from PRBS’s properties. You should be able to get the same results for both of your schemes after disabling.

      +3
    • #21436
      Profile Photo
      Ravil
      Participant

      By the way, I have just check my previous guess with my old version of OptiSystem. I have noticed that each single time (by default) the simulation runs a random value is used for noise and the bit sequence generator (PRBS). Hence, theses random settings should make the resulting BER vary from one run to another! However, if you start increasing bit sequence length the difference between simulations decreases and you receive similar results for all your runs.

    • #21442

      I have the same issue, my colleagues told me one thing can influence is the performance of your computer I mean your ram or even your CPU, cause is because we are using beam propagation method. if you found a way to solve this matter please share, thanks.

    • #21459
      Profile Photo
      Abhishek Shrama
      Participant

      Hello Feras and MohammadMahdi Ariannejad
      the same has been discussed earlier. you can check for details:

      Difference in results in different layouts

    • #21462
      Profile Photo
      Alessandro Festa
      Participant

      If I may add: this should not be surprising. If you run a very short (= short bit sequence) BER test in the lab, you usually end up with different BERs every time…

    • #21466
      Profile Photo
      Ashu verma
      Participant

      Hello, I am agree with Ravil completly.Also Alessandro gave a good point,in general low bit sequence also cause this type of problem.Please visit the link given by abhishek as Damian gave good points to disable the random seed and other parameters to get same results every time from same system.

    • #21609
      Profile Photo
      feras
      Participant

      thank you all , i solved the problem by set the same prbs

      • #21610
        Profile Photo
        Alessandro Festa
        Participant

        You are welcome feras!

      • #21628
        Profile Photo
        Ravil
        Participant

        You are welcome, feras! And thank you for confirmation of my guess. Good luck with your work!

    • #21778
      Profile Photo
      Alessandro Festa
      Participant

      Now that I also discovered this issue, I thinkthat it is better to disable by default the Random Seed noise…I remember in the past many simulations made at small steps of OSNR where the BER did not have a monotonic relation with OSNR and I wondered why!

Login You must be logged in to reply to this topic.