Next Article in Journal
Insights into How Vietnamese Retailers Utilize Social Media to Facilitate Knowledge Creation through the Process of Value Co-Creation
Previous Article in Journal
PVPBC: Privacy and Verifiability Preserving E-Voting Based on Permissioned Blockchain
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Future Internet
Volume 15
Issue 4
10.3390/fi15040122
Review Report
futureinternet-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles
Article
Peer-Review Record

A DNN Architecture Generation Method for DDoS Detection via Genetic Alogrithm

Future Internet 2023, 15(4), 122; https://doi.org/10.3390/fi15040122
by Jiaqi Zhao 1, Ming Xu 1,*, Yunzhi Chen 2 and Guoliang Xu 1
Reviewer 1:
Chin-Shiuh Shieh
Reviewer 2:
Praveen Kumar Donta
Reviewer 3: Anonymous
Future Internet 2023, 15(4), 122; https://doi.org/10.3390/fi15040122
Submission received: 1 March 2023 / Revised: 22 March 2023 / Accepted: 24 March 2023 / Published: 26 March 2023

Round 1

Reviewer 1 Report

Genetic algorithm is employed in this study for the generation of DNN for the detection of DDoS attacks.

Illustrations are in inferior quality.

Existing evolutionary deep learning approaches to DDoD attack detection should be cited and compared.

It is plausible to generate a best fit DNN for a given data set. However, the computational cost needs to be justified. Moreover, it also risks the overfitting issue. The generalization capability of the resulted DNN should be addressed.

Author Response

Dear reviewer, thanks very much for the points you pointed out. Our specific response is attached.

Author Response File: Author Response.docx

Reviewer 2 Report

The following comments need to be addressed before considering this paper for publication.

  1. The abstract and conclusions are very generic. The authors are requested to provide information related to the proposed work, and their pitfalls. Discussing the results from previous methods is also necessary. The conclusion must include one or more future directions. General information such as problem statements and introductions is covered in other parts of the sections, so it is not necessarily included in the abstract and conclusions.

  2. Introduction needs to be improved. Currently, the authors must describe challenges and the need for your contributions. Motivation for choosing this problem is mentioned in its introduction.

  3. Consider state-of-the-art approaches for discussing previous works. It is recommended to summarize the limitations ,of existing works and which among these limitations are addressed in this paper.

  4. This paper uses DNN and GA. There is no clear explanation of each algorithm's purpose and how it addresses DDoS problems. As the algorithm or methodology description seems to be general formulae for DNN and GA, it is very difficult to relate it to the DDoS metrics. So, the authors need to provide a table that maps each metric of DNN or GA with the metrics of DDoS.

  5. Provide the computational efficiency of the proposed work. Compare this efficiency with existing works.

  6. Provide citations for the datasets used in this paper. Considering only one dataset was found in this report, it cannot be used to assess the superiority of the proposed work over the existing work. So, it is recommended to consider multiple datasets to evaluate the performance of the proposed work.

  7. Most of the performance metrics compared in this paper are applicable to DNN and GA. There are no metrics prove DDoS related. In light of the literature that shows several metrics related to DDoS, it is recommended to conduct and present DDoS metrics as well.

  8. The reasons behind the superior performance of the proposed work over the existing ones are not presented in the paper.

  9. A summary of the proposed work's benefits and limitations along with its future scope must be included.

 

Author Response

Dear reviewer, thanks very much for the points you pointed out. Our specific response is attached.

Author Response File: Author Response.docx

Reviewer 3 Report

The application of GA to deep learning and DDoS attacks is interesting. The authors have presented an extensive set of results for three
different data sets. I did not see major issues in the methodology and results.

My main concern is the english writing. The paper is not publishable in this form.

English writing should be improved. There are errors even in the title (Alogrithm) and the keywords.
There are too many errors to report.
I recommend to use a proof-reading service either by MDPI or by some other company.

Other corrections:
DNA(Deoxyribonucleic Acid) is wrong for this context.
Please, increase the FontSize for Figures 1 and 2.
Punctuations problems: there are spaces to be inserted in many places in the
manuscript. Please, correct them.

Author Response

Dear reviewer, thanks very much for the points you pointed out. Our specific response is attached.

Author Response File: Author Response.docx

Round 2

Reviewer 2 Report

The abstract still not informative. The authors are requested to provide the key objectives and metrics of your work in the abstract.

Author Response

Dear reviewer, thanks very much for the points you pointed out. Our specific response is attached.

Author Response File: Author Response.docx

Back to TopTop