AI-Enabled Framework for Mobile Network Experimentation Leveraging ChatGPT: Case Study of Channel Capacity Calculation for η-µ Fading and Co-Channel Interference
Round 1
Reviewer 1 Report
The authors proposed :
REVIEW ON:"AI-enabled Framework for Mobile Network Experimentation Leveraging ChatGPT: Case Study of Channel Capacity Calculation for η-µ Fading and Co-Channel Interference"
However there are a lot of concerns that the paper needs to resolve before it can be reviewed as follows:
1. Lines 52 CCI is caused by various factors as are weather conditions, administrative factors 52
or some design issue-------Grammatical error
2. line 53-54 --What is the meaning of this: The influence of CCI in addition to -μ fading is included in the 53
consideration in [9,10].?
3. Lines 55-56 Grametrtical error on :A standard manner for mitigation the impacts of fading and CCI is diversity combin- 55
ing [1]
4.The paper contribution must be shown and should be clear
5. Paer organization should be consistently clear
6. Authors talked of using this methods to impprove the problem reslution of research so reviwer expects to use that on basis of evaluation of proposed method:
in this paper we will analyze a
multi-branch SC receiver using to improve wireless system in the presence of -μ fading 60
and CCI.
7. What is format1 in :Since η−μ distribution is of Format 1 [3], its parameters are defined as:
8.Line 103 --what is (A5) in The cumulative distribution function (CDF) of SIRs zi are derived in the form of (A5)?
9. Justification of interpretation of equation 7 is not clear basedon lines 106 to 108 explaination
10. The meaning of equation 11 should be clear and well interpreted. Also remove all multiplication signs to the equations not needed.
11. Lines 160 - 161 stament not clear :From Table 1 one sees the number of additions required for reach necessary precision
when changing the parameters μ1 and μ2?
12. Lines 17-172 authors made refrence on I think equation (11) on table 1 based on:
able 1. The number of adders in the sum in expression (11) to achieve precision at the 5th signifi- 165
cant digit-------------Where is Z in the equation 11 on table?
13.L refrred by authors as input branch should be properly defined right from beginning
14. How does Neo4j and the channel fading authors proposed for QoS determined and the Chat GPT is incomplete and i think authors should relate the two experiment well
Thus the machine learning aspect is confusing when linking to
Needs some grammatical construction improve
Author Response
Please see the attachment.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
This paper proposes a classification-based method for quality of service (QoS) prediction and uses ChatGPT and Neo4j graph database to automate the machine learning-assisted experimental workflow in mobile network planning. The experimental results show that the method can effectively save time and cost. But we have some suggestions for this article:
(1) The authors clearly stated in the text that the training and test sets are 80% and 20% respectively. Why are the training set and test set in Table 5 75% and 25%?
(2) Regarding the classification experiment part of this article, I only saw the experimental results of the method proposed in this article and did not see the comparative experiments between this method and other methods, which is not convincing. It is suggested that the author add some comparative experimental results.
(3) According to the experimental results, the authors show that the proposed method can effectively reduce the experimental time cost compared with the traditional method. However, only the time spent on this method is given in the text, and the time spent on the traditional method is not given. It is recommended to add some timetables spent by the traditional method.
(4) Some things could be improved with the picture titles in this article. For example, the title of Figure 3 is placed above the picture, while other picture titles are placed below the picture. It is recommended that the author keep it consistent. In addition, I strongly recommend that authors replace Figure 3 with pseudocode if the algorithm described is really important.
Excepte for some typos and tense erros, I cannot find other errors.
Author Response
Please see the attachment
Author Response File: 
Author Response.docx
Reviewer 3 Report
This paper explores the potential of the new ChatGPT conversational agent in synergy with a model-driven approach based on the Neo4j graph database to aid experimentation and analytics when communication channel performance is considered. The study is limited in terms of novelty. How ChatGPT is used in the study should be shown with examples.
In the abstract and results section, the numerical results obtained should be shared comparatively. Verbal comparison statements such as "The proposed solution reduces the time required for the experiment compared to the traditional approach" are not meaningful and sufficient for making comparisons. The findings (i.e., performance metrics) should also be included in the Abstract and conclusion sections.
The abstract is not well written. The author(s) need to rewrite it and focus on novelty and comparative analysis performed in the paper. (The abstract should explain what the problem is and how it is solved here.
Subject presentation matter should be improved. The motivation for the study is unclear. All uncertainty should be highlighted on the paper.
The fact that the innovative aspect of the study is only "time" when compared to the traditional method is insufficient in terms of contribution to the recent literature. Accuracy performance in parameters such as channel capacity, QoS should also be evaluated.
The experiments and quantitative results are not presented well. The results section must contain quantitative results in the form of a table and these must be discussed in detail to convey to the reader how these results are obtained and why they are valuable.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Reviewer 4 Report
AI-enabled Framework for Mobile Network Experimentation Leveraging ChatGPT: Case Study of Channel Capacity Calculation for η-µ Fading and Co-Channel Interference
Authors: Dragana Krstic, et al
The paper focuses on leveraging AI, particularly ChatGPT, in conjunction with a model-driven approach and Neo4j graph database to enhance experimentation and analysis of communication channel performance in mobile and wireless networks. The integration aims to address factors like fading and interference, while also automating various steps and ultimately reducing the time needed for experimentation.
Overall the paper is well written and a nice contribution to the literature on the use of these technologies to aid experimentation and analytics in the context of communication channel performance, particularly considering factors like fading and cochannel interference and should be interesting for the community who are studying these but needs substantial improvement by providing a more complete exposure to the field. I am happy to recommend the paper to be published in Electronics provided that clarifications made satisfactorily.
Detailed comments on the manuscript are provided below.
SPECIFIC COMMENTS (MAJOR)
1. The introduction is well written but needs expanding to make it clearer to a wider audience. For example, non-markovianity is useful to enhance security in communication, see eg Vasile et al "Continuous variable quantum key distribution in non-Markovian channels", Phys. Rev. A 83, 042321 (2011) and photonic band gap media are promising to obtain nonMarkovian behaviour, Sci Rep 12, 11646 (2022). https://doi.org/10.1038/s41598-022-15865-5,
https://doi.org/10.1016/j.physleta.2022.128022.
2. I think it might be great to think about recent works on applications how phase modulation of coherent states play in quantum communication channels like in the paper: DOI 10.1088/0031-8949/90/7/074027 https://journals.aps.org/pra/abstract/10.1103/PhysRevA.92.012317 and the use of probabilistic noiseless linear amplifiers both at the encoding stage https://doi.org/10.1364/JOSAB.36.002938 where the information is coded on phase shifts and at the decoding stage https://journals.aps.org/pra/abstract/10.1103/PhysRevA.93.062315.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Reviewer 5 Report
This study aims to utilize a recent artificial intelligence tool, ChatGPT, to identify/enhance wireless communication system channel characteristics. Lack of expected rigor and research scope is a main concern. Major observations:
+- Abstract tentatively provides the basics of the research while missing the convincing rationale portion.
-- Introduction is very poor to develop a perspective for artificial intelligence implementations, by using a recent tool of ChatGPT system, in wireless communication systems and related channel modeling as it is the focus of the study (while giving traditional wireless communication system channel related perspectives).
- Define every parameter of an equation as most are missing in the text.
- Equation 11 seems odd to effectively handle the channel capacity and focus on a particular aspect.
- Section 2 is mostly procedural and borrows from the literature. However, selection of the parameters, physical implementation and interpretation considerations, illustrative powers of the simulated channels, etc. are all missing.
-- Section 3 Neo4j database management system needs to be convincingly elaborated to demonstrate the potential in wireless communication channel considerations. Also, the related model training algorithm (borrowed from an existing study) elaborations and related justifications for the underlying research are completely missing. Whole Section 3 narrative sounds more of a software tool implementation procedure/study with inconclusive outcomes, than a rigorous research integration of artificial intelligence in wireless communication channels.
- Table 3 needs to be revised to focus on its content flow, expected information processing, potential superior operations, etc.
- "Experiments" referrals need to be substituted by "simulations".
- Many noticeable presentation issues imply a thorough content and format edition: - Unexpected reference in Abstract, - Incomplete and noncohesive paragraphs, e.g., line-36, - Lumped references need to be expanded concisely and linked to the underlying study, e.g., line-42, - Improper sentences are observed, e.g., line-45, - "All off them ", - "In this figure ", - "Below we give two tables ", - One-sentence paragraphs, e.g., line-157, - Improper caption placements, e.g., line-208, - "mesa-otpimization ", - "model the conforms ", - Table 4 is very poorly presented.
Please see above relevant comments.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Round 2
Reviewer 1 Report
Authors proposed:
"AI-enabled Framework for Mobile Network Experimentation Leveraging ChatGPT: Case Study of Channel Capacity Calculation for η-µ Fading and Co-Channel Interference"
The authors have modified the paper and explained methods clearly.
Per can be accepted after minor editing;
The multiplication sign at the beginning of equations starting form the lines 830-862 is uneccessary same applies to the appendix section which should be resolved
Minor english editing in whole article structure should be resolved before paper publication
Author Response
The multiplication sign at the beginning of equations starting form the lines 830-862 is unnecessary; same applies to the appendix section which should be resolved
Our answer:
Thank you for repeated remark. Sorry because we did not correct this previous time.
Now, this issue is corrected in the new version of the paper. We removed the multiplication signs in every line where they appeared unnecessarily.
Reviewer 2 Report
The authors modified this article according to our previous suggestions. But there are still some issues that need the authors’ attention:
(1) Line 54: when an English abbreviation (MRC) is used for the first time, the complete vocabulary should be added to facilitate understanding.
(2) Line 121: there is a syntax error, “at the for”.
(3) In Section 3.5, it is suggested to add some pictures about the result data to make it more convincing, instead of just listing the results through a table. Such as adding pictures of the training set and test set results of the classification model, etc.
(4) There are some mistakes in the presentation of the paper, which make it difficult for readers to understand
There are some mistakes in the presentation of the paper, which make it difficult for readers to understand
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Reviewer 3 Report
The authors have progressed in improving the paper (electronics-2570235-peer-review-v2) compared to previous versions of the paper (electronics-2570235-peer-review-v1). When both versions are evaluated comparatively, it is seen that the authors make the corrections requested by the referees and show the necessary sensitivity in the revision of the paper in line with the comments. In the revised version of the paper, almost all the comments have been considered and addressed by the authors. The response to reviewers file is well-prepared. The changes/corrections made by the authors in line with the opinions/suggestions/evaluations of the referees can be tracked easily. The existing organization and spelling problems in the previous version of the paper have been fixed.
I would like to thank the authors for their efforts for the edits/corrections requested by the reviewers and for the well-prepared response file.
As a result, my concerns on the previous version of the paper have disappeared with the explanations made by the authors, as well as the revision they have made.
This revision is sufficient, and it is possible to evaluate the paper for publication after preparation according to MDPI Journal template.
Author Response
Thank you once again for useful comments and remarks. We gave our best to increase the quality of our paper as much as possible.
Reviewer 4 Report
Dear Authors,
The new version of the manuscript addresses most of the worries and I am therefore recommending acceptance of the manuscript.
Author Response
Thank you for your useful suggestions in the previous round, which we seriously took into account in order to improve the overall quality of the paper.
Round 3
Reviewer 2 Report
Many problems pointed out before still exist, and I regret that I cannot recommend this paper for publication.
There are tense and grammar issues in the paper
Author Response
Dear reviewer,
We gave our best to proofread the text several more times involving both the co-authors and external persons highly proficient in English in order to improve the quality regarding both the language-related aspects, such as grammar, fixing typos and tenses, together with phrases and overall writing style contributing to paper’s presentation quality. When it comes to other modifications from previous rounds, we already covered as much as possible all the missing elements and corrected the noticed issues, giving our best to exhaustively address the remarks noticed by all the reviewers involved.
We double checked all the remarks from past rounds and strongly believe that the current version is aligned with all the reviewers’ comments. Additionally, we also gave our best to identify other issues not identified by reviewers in order to provide the highest possible quality of our submission.
Additionally, we restructured and shortened many sentences in the introduction in order to improve the presentation quality of this paper, making its structure and contribution overview more clear. We hope that the latest corrections of the English in the paper will finally highlight the scientific quality and contributions it has.
With respect,
Authors
Author Response File: Author Response.pdf
Round 4
Reviewer 2 Report
All my questions have been addressed, I have no more questions
There are some tense errors.
