Machine Learning Based Interference Mitigation for Intelligent Air-to-Ground Internet of Things

Round 1

Reviewer 1 Report

The paper "Machine Learning Based Interference Mitigation for Intelligent ATG IoT" is within the scope fo Electronics. The authors investigate the  interference caused by the frequency reuse between ATG system and ground mobile communication system. In particular, they analyzed the interference from 5G ATG airborne equipment to ground 5G system in 3.5 GHz frequency band.

Please avoid the use of acronyms in the title. It is not a scientific approach, although the reader knows the meaning.

The abstract should be revised. It is too short, it does not clarify the content of the paper. The results are not presented, the methods are not presented, the content is too vague. Moreover, the content of the abstract is confused with the content of the introduction. It is suggested to refine the abstract and highlight the key work of this paper.

The review literature should be improved.  The bibliography is too directed towards authors of Chinese nationality. Not that this is an absolute problem, but the contribution of previous knowledge to the research presented should be balanced.

Equations should be numbered. All the variables should be defined.

The methodologies are not clearly described, the results should be commented. Some figures are not discussed and explained.

 

There are several typing errors.

The English language should be revised.

 

 

Author Response

The responses to the review comments are given in the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

The manuscript entitled “Machine Learning Based Interference Mitigation for Intelligent 2 ATG IoT” is submitted for possible publication in Electronics. The paper sets out to investigate co-channel interference between ATG and 5G at 3.5 GHz. The paper has some useful information. Major revisions are however needed.

Comments

1.      The Abstract section is poorly structured. It does not provide the needed information to the reader. The author (s) should re-write this section.

2.      The specific objectives and gap(s) in the literature are not clearly stated. Furthermore, the novel part of the paper is missing from the Introduction section.

3.      Author(s) should state the rationale for adopting the prediction method of time series based on ML. There are other methods available in the literature. I think the author(s) should add more text here to talk about the wide-ranging methods.

4.      In equation 8, the author(s) need to address the importance of the scaling factor. Also, any assumptions impeded in the formulation of equation 8 should be spelled out.

5.      Model validation is missing – how can we judge the estimated ML model?

6.      The Conclusion section is not well-structured – a great deal of repetition. Further, I do not see solid recommendations.    

 

7.      I think the limitation(s) of the presented method should be covered in a comprehensive manner.

Author Response

The responses to the review comments are given in the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

It is not an excellent paper. The conclusions still should be improved.

Author Response

The responses to the review comments are given in the attachment, thanks.

Author Response File: Author Response.pdf

Reviewer 2 Report

The reviewer has reviewed the revised version of the paper. Most of the comments are addressed. I have no further comments. 

Author Response

Thank you very much for your valuable comments on the improvement of our paper.