Tunable Optoelectronic Chromatic Dispersion Compensation Based on Machine Learning for Short-Reach Transmission
Round 1
Reviewer 1 Report
This work is an extension of an already published work, with crystal clear and substantial added value. The results are clearly justified. The paper is very well structured. Therefore I consider it as an accepted research work and hence, I recommend it for publication.
However, the writing is quite bad in a few sentences, mostly the syntax is not proper, thus making it hard for a reader to comprehend the meaning of the sentence. Below I am referring to the specific sentences that must be rewritten with proper syntax.
Lines 36-37 sentence “However..”
Lines 107-108 sentence “Considering..”
Lines 117-118 sentence “The necessary..”
Line 118 sentence “An addition..”
Lines 157-159 sentence “The difference..”
Lines 175-176 sentence “We note..”
Line 195 sentence “Fig.9..”
Also some minor comments
Line 3 use “is” instead of “are”
Line 14 use “are” instead of “is”
Lines 22-24 a reference is required for sentence “Alternatively, solutions…”
Legend of Fig.1 and specifically the part “Different receivers were analyzed and are shown in Fig. 2”. I don’t think it is proper for a Fig. legend.
Line 117 use “connected” instead of “connect”
Lines 137-138 a reference is required for sentence “The MLSE.…”
Line 141 a reference is required for KP4 FEC
Line 142 what’s the meaning of number “2” in the beginning of the sentence?
And finally a couple of generic questions
Line 58 a root-raised cosine (RRC) is used as pulse shaping. Why the signal has to be changed to analog instead of remaining constantly digital through all the root form transmitter to receiver?
Line 69 “MZM is not simulated”. Aren’t we expecting any extra signal deterioration from MZM? By omitting the modulator isn’t it somehow a more ideal scenario?
Lines 195-196 sentence “First we notice…” How is it explained that similarity?
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Reviewer 2 Report
A hybrid optical and digital ML based CD compensation method is proposed in this paper. This paper is written well, but it may hide some key issues in realistic inter-data center networks. I believe DTU has the ability to experimentally test this scheme. When we try to resolve the commercial inter-data center related issues, I think experiment is very important. My comments can be found below.
References [5-9] discuss the other ways to mitigate CD. But it is not the whole picture. There are some papers published in ECOC and OFC about DD based CD mitigation, some of which can have great performance and complexity. The authors should study these idea, and compare your methods with theirs. I haven't seen large Baud rate C-band OOK in pluggable optics market. But PAM4 may be viable. Is the proposed method suitable for PAM4 based C-band LR range transmission? MZM is polarization sensitive, but the polarization of the signal is always changing during fiber transmission. How to align the MZDI with the ever-changing polarization. What is the sampling rate requirement of the ADC for the hybrid CD mitigation method?Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Round 2
Reviewer 2 Report
The authors addressed my comments well. I have no further comment. My only concern is that this scheme may not suitable to be used in DC networks. At least the authors have to solve the MZM polarization dependent problem. I guess that is the main reason why this is a numerical simulation paper.
Author Response
Dear reviewer,
Thank you for your concern about our next step, the experimental setup. We think the use of MZM will be a challenge for the reasons you mention in the review and because of that, we will also consider the use of an AWG, instead.
We appreciate the comments of the reviewers in this publication and the concern of the next steps. The manuscript improved compared to the original version.
Yours sincerely,
Stenio Magalhães Ranzini