Different results from Optiwave V. 7 & v.13

[ahmed helmy teama]

Dear all,
i ran same system as attached in Optiwave V.7& V.13 , it gives me different results
V. 7 gives me Qfactor=8.23
V. 13 gives me Qfactor=4.7
Anyone can help me?

[ahmed helmy teama]

Sorry the attached file missing

Attachments:

1. fso1.osd

[alistu]

Hi Ahmed, I ran the file and it was interrupted by an error so I was forced to close the program! I think the differences between the results obtained in obviously due to the differences in how the component models have changed for the good, and you can definitely rely on the results obtained by the newer versions more than those of the older ones.

[Ashu verma]

Hello Ahmed
As Alisthu said about error in this system file.But Results not have to be vary for same file in two different versions.There are two different aspects, first is if you are getting different results every time from same system in one particular version and other one is by changing the version u sing same file.For first ,you have to disable random seed generation from some components.For second one you have to contact optiwave team i think. Thank you

[alistu]

Hi Sam Sung, I believe getting different answers every time you run the simulation seems to be a very serious problem. By enabling random seed generation just minor differences are expected to be observed in the results. I always enable random seed generation, yet I always get results more or less the same.

[Ashu verma]

In one of my system in optisystem older version i was getting the varied results every time i ran the simulation.And the difference is of 5 to in Q factor values .One time it is 6 Q-factor other time 10.Damian suggested me to deselect the random seed noise.Also i was asked to disable the noised like ASE ASE-ASE fro photodetector.

[K. Esakki Muthu]

it is not about random seed, Still there is a problem, i have experienced the same few months back.? I don’t know the reason behind this.Kindly anyone please answer this.

[alistu]

Hi Esakkimuthu, The problem you are referring to comes really strange to me. In the case Sam Sung mentioned, minor differences are possible but not in yours. Since I have never had this problem, may I ask you to attach a file for whose implemented system the aforementioned problem happens? Thank you.

[Damian Marek]

It is hard to say what is happening here, because version 7 is quite old. Depending on how you built the project could lead to different results because some layout parameters were brought into newer versions. It could also be the random seed, but that seems like a large difference to attribute to randomness. Have you tried increasing the sequence length to see if that levels out the difference in Q factor?

[ahmed helmy teama]

Dear Damian,
I changed the sequence length from 164 bits to 256, buy there is no change in results in both versions.

[alistu]

Hi Ahmed, Based on what Damian mentioned yesterday in another topic I think if the problem is actually from the sequence length, then you must increase it far more than that. I suggest that you use a sequence of length 4096 and see if the results change.

Cheers!

[Author]

Posts