Dear Everyone, I am presently Working on WDM system Design Using Optisystem V13 Simulation platform. I am finding some kind of unusual Eye Diagram. Can anyone tell me What may be the probable reason of not clear Eye diagram as in the attached image. Thanking you.
Regads,
Dhiman
Difficult to be certain without more information.
How are you modulating the transmitter, what is the bit sequence and what kind of receiver are you using?
The signal is bipolar, so you are not plotting intensity directly. It is presumably either a direct visualisation of the optical field, or the output of an AC-coupled or differential coherent receiver.
It could be some kind of 3-level modulation such as a phase duobinary variant, but you would presumably know if that were the case. More likely is QPSK modulation with a 45 degree phase shift in the local oscillator compared with an ideal configuration able to deliver independent I and Q channels via a simple binary decision threshold. The phase shift is not a concern if you have an ADC on each of two quadrature channels and rely on DSP for phase recovery, but it does make it more difficult to analyse what is going on.
The signal changes state every symbol, so you could be using a sparse, very short or non-random bit sequence, or a coding scheme which guarantees a transition every symbol.
Any of this of help?
Sir,
The following are the specification I am using,
Modulation Format at the transmitter: DQPSK
Bit rate: 10Gbps
Sequence length: 128bits
receiver: Direct detection
Polarity: Unipolar
Sir I have attache here with the screenshot, Please have a look.
Thanking you,
With regards,
Dhiman Kakati
Dhiman,
I still don't have a clear view of what is happening, but I would question the delay of 2 bit periods in the receiver Mach-Zehnder demodulator. In a conventional DQPSK scheme, a delay of 1 symbol is required.
This assumes that the "fork" and "4-DPSK precoder" blocks do what I expect.
There is confusion in the terminology used in the summary table - possibly because Optisystem is trying to maintain backward compatibility using names derived for binary modulation, which are ambiguous or incorrect when applied to multi-level modulation formats.
As far as I can tell, you sample the waveform at 2.56 THz, with 256 samples in each of 128 symbols for a sequence length of 12.8 ns. The symbol rate is 10 GBaud, with 2 symbols/bit for an aggregate bit rate of 20 Gbit/s.
Alternatively, is the symbol rate 5 GBaud with 64 symbols each of 200 ps period? In this case your demodulator MZ delay of 2/Bit rate is correct.
The summary table specifies a symbol rate of 2.5 GBaud, which is inconsistent with both interpretations.
The pulse carver uses a sine wave driven Mach-Zehnder at 1/4 Bit rate. Is this 2.5 GHz? Is the pulse-carver Mach-Zehnder biased to produce a 2.5 GHz or a 5 GHz pulse train? A 2.5 GHz pulse train would explain the zeros in the eye.
Can you clarify what your symbol rate and symbol period is? Are you able to display the pulse carver and signal waveforms or the eye diagram with an absolute time axis (ps or ns) to resolve this ambiguity?
Regards,
Alan.
Thank you so much sir for your kind support, taking consideration of you question Now I got the output. I am attaching herewith the snapshot, I think this is acceptable.
With Regards,
Dhiman Kakati
- Badges
- Science topic
- Similar topics
- Mathematical Sciences
- Graphs