Detecting DDoS Attacks in IoT-Based Networks Using Matrix Profile

Round 1

Reviewer 1 Report

Matrix profile is employed in this study of DDoS detection in IoT networks.

The experiment demonstrates the feasibility of the proposed approach.

It appears the matrix profile approach is computation intensive. Will it be feasible for IoT devices with limited computing power? In particular, the experiment was conducted on a PC platform.

The nature of the data set demands further elaboration.

Is real-time detection possible? No computing time is reported in the paper.

The lack of comparative study renders this study less persuasive.

Author Response

Please see the attachment

Author Response File: Author Response.docx

Reviewer 2 Report

Dear authors thank you for the interesting article. The text is well structured, and the description explains the author's idea to the bottom. Is worth mentioning that the article consists of a complete review of similar research papers.

Before publication, please make the following changes to the text:

- in the figure number 3 description, consider adding the text „axis” after x, y 

- currently, the graphs in figure 15 are unreadable, please change their layout or show different graphs on the same graph, using color to differentiate the time series,

- please consider reformulating your comments about other articles, in lines 33-34, 87-89, 175-178,

- line 298, add text “figure” before 2 and 3 numbers,

- in line 222 authors mentioned computation time, please add a few sentences in the conclusions section, describing the computation time of the author's method, and the possibility of implementation in real-time, maybe. Currently, we have here only information about algorithm precision. 

Author Response

 

1. in the figure number 3 description, consider adding the text „axis” after x, y.  Done 
2. currently, the graphs in figure 15 are unreadable, please change their layout or show different graphs on the same graph, using color to differentiate the time series.  Done 
3. please consider reformulating your comments about other articles, in lines 33-34, 87-89, 175-178,  Done 
4. line 298, add text “figure” before 2 and 3 numbers, Done 
5. in line 222 authors mentioned computation time, please add a few sentences in the conclusions section, describing the computation time of the author's method, and the possibility of implementation in real-time, maybe. Currently, we have here only information about algorithm precision.  Done 

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Issues had been addressed, although not perfectly.