[mai fouadParticipant]

attached here a file comparing the performance of DSP and DCF, the difference in the performance can be seen from the constellation visualizer.
i don’t know where is the problem, the frequency domain filter just multiply the signal by the conjugate of {exp(j*[(beta2*omega^2/2)+(beta3*omega^3/6)]), beta2,3: second and third order dispersion, omega: angular frequency} to compensate dispersion.

Attachments:

1. 4_QAM_dispersion_compensation.osd

[mai fouadParticipant]

I attached these files, that load In-phase and q-phase components from Matlab to optisystem also the binary data, QPSK symbols for calculating BER if needed and you may need (Inphase_comp_tx_for_Q-factor)that contains In-phase component with no biasing or amplification (i.e. constellation points at [-1+1j 1+1j -1-1j 1-1j]). you could use these files and change the paths or use your own QPSK TX and RX but you will need the last two matlab component(specially matlab component 2) to compensate the dispersion.

Attachments:

1. data-files.zip

[mai fouadParticipant]

I agree with you that the problem in the noise of the system, as when i test LMS algorithm by just adding noise to the modulated signal (without optical fiber channel), it couldn’t compensate its effect, does this problem usually occur with LMS algorithm?

Attachments:

1. LMS.zip

[mai fouadParticipant]

also i tried to simulate DP-QPSK but the problem still exists

Attachments:

1. dual-polarization.osd

[mai fouadParticipant]

i catch the error , it was at transmitter that i used CW laser instead of spatial CW laser .

[mai fouadParticipant]

I found this problem when i import my index profile either in .txt or .dat , it give me error in loading file , file is corrupted : here is the error and profile

Attachments:

1. 

   optifiber-error1.png
   

1. design.rar

[mai fouadParticipant]

i think that there is an error in loading the index profile txt file , but how to solve it ?

[mai fouadParticipant]

i see that but these parameters control the dispersion in optical fiber , i asked about the linear refractive index of the fiber can be changed in this component or not !!!

[mai fouadParticipant]

but DGD differential group delay related to delay between two polarization modes not delay between propagation time between two different fibers !!!

[mai fouadParticipant]

but this matlab component followed by another components in the system and i want to calculate BER at each time of running (i.e i want to take a vector containing all BER calculated for each run )

[mai fouadParticipant]

here is my project and the m-file called in the project but it will take long time to run also if i try another project and put the sequence length as i mentioned i get the same error

Attachments:

1. try1.osd
2. TEST.m

[mai fouadParticipant]

yes i use sequence length =1048576 bits as my system has a small BER around 10^-5 so i need to test the system with million bits , is there another method to test huge number of bits in optisystem