A Channel Model to Deal with Distributed Noises and Nonlinear Effects in a Fiber System with Distributed Raman Amplifiers
Round 1
Reviewer 1 Report
The authors propose a new model to describe fiber systems with distributed Raman scattering.
This paper is nicely presented. It includes a comparison with the original Schrodinger equation approach and experimental results. The language and the structure are clear.
I recommend acceptance.
Author Response
Thank you very much for your attention and the referees’ comments on our paper "A Channel Model to Deal with Distributed Noises and Nonlinear Effects in the Fiber System with Distributed Raman Amplifiers ".
Reviewer 2 Report
This paper should show the important contributions for the proposed numerical analysis and experiments.
Author Response
Thank you very much for your attention and the referees’ comments on our paper "A Channel Model to Deal with Distributed Noises and Nonlinear Effects in the Fiber System with Distributed Raman Amplifiers ".
Reviewer 3 Report
This paper describes a Channel Model to Deal with Distributed Noises 2 and Nonlinear Effects in the Fiber System with 3 Distributed Raman Amplifiers. The paper needs a minor revision because it needs to separate the methods from Results and Discussion. The way it is presented it is very difficult to overall understand what are the methods and which are the results!!
Author Response
Dear Reviewer,
Thank you very much for your attention and the referees’ comments on our paper "A Channel Model to Deal with Distributed Noises and Nonlinear Effects in the Fiber System with Distributed Raman Amplifiers ".
We have revised the manuscript according to your kind advice.
Thank you very much for all your help and looking forward to hearing from you soon.
Comment: This paper describes a Channel Model to Deal with Distributed Noises and Nonlinear Effects in the Fiber System with Distributed Raman Amplifiers. The paper needs a minor revision because it needs to separate the methods from Results and Discussion. The way it is presented it is very difficult to overall understand what are the methods and which are the results!! Response to the reviewer:
Thank you for your comments and advice!
We feel very grateful for your advice to separate the methods from Results and Discussion. In the revised manuscript, we added a chapter named “Discussion” to analyze the results. And we change the chapater “Experiments and Analysis” to “Experiments”.
Line 253 to Line 275:
DiscussionIn the experiments, we compared the simulation results of our model and the simplified NLSE with the experimental results of the fiber systems with only backward-pumped FRA (System A), both forward- and backward-pumped FRA (System B), and 6-span 80km fiber with a 16dB-gain back-pumped FRA (System C).
In System A, there was only a bit of difference between the two models. When the optical launch power is higher than 9dBm, the EVM in our model increases a little sharper than that of simplified NLSE. It might be caused by the influence of the RIN transfer between the pump light and the signal light. In simplified NLSE, all noise was regarded as the centralized additive white Gaussian noise (AWGN) and was added to the signal field at the receiver. However, the RIN impacted with the signal distributedly in our model, which was more sensitive to the fiber transmission. Since there was little difference in the errors of our model and the simplified NLSE with the experiments, we could not tell which model was better.
For System B, in the system with forward-pumped FRA, the RIN and SpRS noise were more severe. Since the signal power was high, the distributed noise could interact with the signal for the nonlinear effects. Therefore, it was improper to regard the distributed noise as centralized AWGN at the high launch power.
In System C, the simulating results of our model were fit better with the experimental measurements, which verified our model was effective to describe the multi-span system with backward-pumped FRA.
In summary, our modified model has the bttter performance for the simulation of the fiber system with forward FRA and high nonlinear effects. And the simplified NLSE model is suited to the conventional fiber systems with EDFA and the single-span fiber system with only backward FRA.
Author Response File: 
Author Response.docx
Reviewer 4 Report
Please see attached file.
Comments for author File: 
Comments.pdf
Author Response
Dear Reviewer,
Thank you very much for your attention and the referees’ comments on our paper "A Channel Model to Deal with Distributed Noises and Nonlinear Effects in the Fiber System with Distributed Raman Amplifiers ".
We have revised the manuscript according to your kind advice.
Thank you very much for all your help and looking forward to hearing from you soon.
Comment:In this manuscript, the authors propose a modified nonlinear Schrodinger equation (NLSE), which takes into account most of the physical processes of distributed fiber Raman amplifier (FRA). This method has better capability to model the distributed noise than simplified NLSE and takes “acceptable computation cost”.
The presentation of this paper is largely easy to follow. The introductions and motivations are adequate. The main results are clearly presented with enough context. Given that this is well structured draft with sound results and conclusions, the English language clearly needs improvement. There are many gramma mistakes and misnomers within the draft, which makes it difficult for readers to fully appreciate the significance of this paper. I strongly recommend the authors to double check the language and, if needed, find some external help about the English.
Response to the reviewer:Thank you for your comment!
We have checked the language and revised the manuscript.
Some other comments:
Comment: On line 96, “Eq. (4)” probably should be “Eq. (3)”. Response to the reviewer:Thank you for your comment!
In the revised manuscript, we have revised the “Eq. (4)” to “Eq. (3)”.
On Line 96 of the revised manuscript:
However, the accuracy of the model based on Eq. (3)is unsatisfied to simulate the performance of FRA system.
Comment: In figure 3, the authors compare simulations from their approach and simplified NLSE to experiments. However, just by looking at the traces, it is not very clear how much improvement the proposed method achieved. If some quantitative measure such RMS can be given, the improvement from using modified NLSE would be more obvious. Response to the reviewer:
Thank you for your comment and advice!
We have given the root mean squared error between the simulation results and experiments in the revised manuscript.
On Line 221 to 223:
The root mean squared error (RMSE) between the EVM of the simplified NLSE and the experimental results is 0.490%; The RMSE between the EVM of our model and the experimental results is 0.431%.
On Line 232 to 234:
The RMSE between the EVM of the simplified NLSE and the experimental results is 0.889%; The RMSE between the EVM of our model and the experimental results is 0.492%.
On Line 250 to 252:
The RMSE between the EVM of the simplified NLSE and the experimental results is 7.396%; The RMSE between the EVM of our model and the experimental results is 1.299%.
Comment: Since the authors claim their modified NLSE offers better simulation while takes “acceptable computation cost”, the estimated computational cost should be given and compared to other models. Response to the reviewer:
Thank you for your comment and advice!
We have given the estimated computational cost of our model in the revised manuscript and compared the computational cost of our model with that of simplified NLSE.
On Line 180 to 186:
In the simulations, we employed the split-step Fourier method (SSFM) to simulate the performance of the fiber system and the number of the steps was 1,000 all the time. The computation cost of the SSFM was from the FFT/IFFT operation. With the same number of steps, both our model and the simplified NLSE need to do 1,000 pairs FFT/IFFT operations. The computation cost of our model and the simplified NLSE in SSFM were almost the same. The increased computation was from the noise calculation by Eq.(7) to Eq.(12). Therefore, the computation cost was acceptable.
Comment: It would be nice if the authors could specify the conditions where the simplified NLSE applies and where only the modified NLSE they developed can give accurate results. Response to the reviewer:
Thank you for your comment and advice!
In the revised manuscript, we have explained the application range of the simplified and modified NLSE.
On Line 273 to 275:
In summary, our modified model has the bttter performance for the simulation of the fiber system with forward FRA and high nonlinear effects. And the simplified NLSE model is suited to the conventional fiber systems with EDFA and the single-span fiber system with only backward FRA.
Author Response File: Author Response.docx
Round 2
Reviewer 2 Report
The simple NLSE model can be used for the fiber system with EDFA which can be applied to more deeply analyzed fiber-amplified system.
