Intelligent Matching of the Control Voltage of Delay Line Interferometers for Differential Phase Shift Keying-Modulated Optical Signals

Round 1

Reviewer 1 Report

The paper is aimed to propose a new metod to fast locate the optimal DLI control voltage  based on FNN. The article is interesting and this topic is relevant.

The article can be accepted for publication after the next edits:

• In my opinion, It is worth specifying the annotation: the problem; what the authors suggest; what results they obtained.
• It is necessary to change the structure of the article a little, to add the Relation works section
• In my opinion, As the topic is new, the authors should add 5-7 publications for 2019-2021.
• Justify the choice of the number of layers of the ANN more precisely.
• To enhance the work, the authors should add the comparisons with experimental results.
• It would be worth to note the measurement error.
• It is desirable to expand the scope of the research and consider in more detail the future of the work.

Author Response

Dear Reviewer,

Thanks for your hard working on our submitted manuscript!

According to your comments and suggestions, we revised the manuscript as follows.

Comment 1:

      In my opinion, It is worth specifying the annotation: the problem; what the authors suggest; what results they obtained

Answer 1:

      We have modified the organization of the manuscript to show the problem, what we suggest and results we obtained more clearly.

Comment 2:

      It is necessary to change the structure of the article a little, to add the Relation works section.

Answer 2:

      We added the relation works section in the revised manuscript.

Comment 3:

      In my opinion, As the topic is new, the authors should add 5-7 publications for 2019-2021

Answer 3:

      We added 6 publications for 2019-2021 to support the use of FNN in optical communication networks for physical layer solutions

Comment 4:

      To enhance the work, the authors should add the comparisons with experimental results

Answer 4:

      We added more comparisons with experimental results in the revised manuscript.

Comment 5:

      It would be worth to note the measurement error

Answer 5:

      We mentioned the measurement error in the revised manuscript.

Comment 6:

      It is desirable to expand the scope of the research and consider in more detail the future of the work

Answer 6:

      We expanded the scope of the research and considered in more details of the future of the work in the Introduction section and discussion section.

Best regards!

Reviewer 2 Report

The authors have presented an interesting work on the use of FNN to speed-up the demodulation process. The overall presentation is good and technically sound. The write-up can be further improved.

In terms of technical revisions, here are my comments

• Please elaborate the function of sinusoid component in Figure 1. How does a sinusoid is used to carve pulses? What should be its frequency then? What if a signal with two times the required frequency is used for this component? for line 118: please explain 33% term. it is not enough to just provide the reference of the papers since this paramter is important to this article. it is somewhat explaiend in equation 13 but still need more clarification and general concept in the paper.
• Line 101: if the MZM is biased at Vpi, then how this non-linear operation affects the overall system? why not quadrature bias point?
• The discussion and conclusions part is not sufficient enough to explain the findings. The authors should revise this section.
• Please replace the symbol for Fiber in Figure 1. Fiber is a passive component and the two arrows directed outwards may not demonstrate the function of a fiber
• Line 85-86, the second MZM (represented as MZM-1 in figure 1). Please check if it should be MZM-2?
• More references are required to support the use of FNN in optical communication networks for physical layer solutions.

Author Response

Dear Reviewer,

Thanks for your hard working on our submitted manuscript!

According to your comments and suggestions, we revised the manuscript as follows.

Comment 1:

      Please elaborate the function of sinusoid component in Figure 1. How does a sinusoid is used to carve pulses? What should be its frequency then? What if a signal with two times the required frequency is used for this component? for line 118: please explain 33% term. it is not enough to just provide the reference of the papers since this paramter is important to this article. it is somewhat explaiend in equation 13 but still need more clarification and general concept in the paper.

Answer 1:

      We added the brief introduction of the function of sinusoid component and the frequency of the sinusoid wave in Figure 1. However, what needs to be pointed out is that our major work is to improve the demodulation method of DPSK modulated optical signals. The modulation procedure of a DPSK signal is a classical technique and is clarified in lots of classical literatures. The modulation procedure is not our work and is not the main contribution of our manuscript. We just refer and using the corresponding basic theories to support our derivations on the design of demodulators at the receiver. The details of DPSK modulations are detailed in the references [25-29] of the revised manuscript.

Comment 2:

      Line 101: if the MZM is biased at Vpi, then how this non-linear operation affects the overall system? why not quadrature bias point?

Answer 2:

      We have revised the manuscript to descript the problem more clearly. The operation referred by the reviewer is not the major reasons of the non-linear effects. This operation is a fixed operation. The non-linear effects are mainly brought in by the machine errors, which affects the phase-error and not easy to estimate, and for different individual DLI devices, it is not the same. So it is difficult to resolve by traditional methods.

Comment 3:

      The discussion and conclusions part is not sufficient enough to explain the findings. The authors should revise this section.

Answer 3:

      To enhance the work, we added more comparisons with experimental results in the revised manuscript and present more details in the discussion section.

Comment 4:

        Please replace the symbol for Fiber in Figure 1. Fiber is a passive component and the two arrows directed outwards may not demonstrate the function of a fiber

Answer 4:

        We modified this error in the revised manuscript.

Comment 5:

        Line 85-86, the second MZM (represented as MZM-1 in figure 1). Please check if it should be MZM-2?

Answer 5:

        We modified this error in the revised manuscript.

Comment 6:

      More references are required to support the use of FNN in optical communication networks for physical layer solutions

Answer 6:

      We added 6 publications for 2019-2021 to support the use of FNN in optical communication networks for physical layer solutions

Best regards!