Dynamic Offloading Model for Distributed Collaboration in Edge Computing: A Use Case on Forest Fires Management

Round 1

Reviewer 1 Report

This paper is well written but not well structured.
The example of the fires should be shown even if it is a simulation. The concrete study should be completed, it is an applied journal.

Author Response

This paper is well written but not well structured.

-> The paper changed the structure to fit the journal.

The example of the fires should be shown even if it is a simulation. The concrete study should be completed, it is an applied journal.

-> This paper is a model-proposed paper that was written for a modeling scenario rather than an actual experiment. In the next paper, we will add the experiment based on this content. 

Author Response File: Author Response.docx

Reviewer 2 Report

Nice original work. This work is well studied especially the limitations of other offloading based on edge computing. However, there is need that the authors show the reads how their dynamic offloading is better especially in fire prediction to support prevention. The structure of the paper needs to be improved. the introduction is lacking wildfires information and yet its based on wildfires that you want to the proof that your dynamic offloading model is better than the others. The methods have been mixed with results and discussion. The discussion section is needed for comparison with other scientific studies already done. I also encourage the authors to present the Figures 1,2,3 and 4 as results and later discuss them in the discussion section. Also show at least one out put from your dynamic offloading model on fires because i did not see any.

Comments to authors

The paper is a good but revise the comments bellow.

line 18 change Edge Computing to Edge computing in the whole manuscript.

line 32-34 does not make sense. I think you should insert "is not only using numerical data...." after the word analysis so that the sentence makes sense to the reader. 

line 36 remove the comma after latency and do the same in the whole manuscript. 

line 37 put the full stop after citation [3-7] and [8-13] remember to do the same in the whole manuscript.

line 42 make the citations this way as required [15-17] and remember to do it in the whole manuscript.

line 44 remove "However" at the start of the sentence and remove "now" after cloud computing

line 45 begin the line with "However,...........................

line 48 change the sentence Other Edge computing to issues to consider are response time requirement,.........,.........,..remember to change the citations to [18,19].

Line 50 does not make sense .......Begin with Edge computing main offloading task is dividing tasks that require.................full stop. Then begin As a result, there are offloading........

lines 53 Begin this way.... This load is a problem in areas where immediate response is needed to detect nearby objects, collect real time data and speed up complex calculations [20-22]. Then move lines 61-64 here.

Then remove example 1 and 2 and focus on fire because your paper applies a model on fire. Just begin For example, if fire is detected...............In recent years, the seriousness of wildfires is rapidly increasing. Then continue and write one or two paragraphs explaining why the other offloading approaches are not good for forest fires or wildfires and why you propose the use of algorithm-based offloading on a risky situation that needs rapid decision action like forest fires.

line 66 remove where and replace with that

line 70 change was and replace with attempt has been made........

line 72 change to Section 3 includes an .........

line 74. Change to The conclusions of this paper are in section 5.

lines 76 should read Methods

lines 77-91 should be moved to introduction to explain why other data or task offloading have failed to address forest or wildfires

line 92  should read, We propose a model.......... and delete line 93 sentence Chapter 2.1 describes edge computing based offloading in line detail

The methods section is not consistent.

I think lines 167-175 should be in the methods section. As the explain the edge node

I suggest that they reconsider the methods section as follows

2.1 Edge note

2.2 Master Node

2.2.1 Data centric offloading algorithm

2.2.2 Task centric offloading algorithm

2.3 Second master node

2.4 Action Node

2.5 Forest fire prediction as shown in lines 257-295

3.0 Results

3.1 Offloading based on edge computing...................As shown in lines 94-103

Lines 104-121 move to discussion section

3.2 Problems of offloading based on edge computing As shown in lines 125-129

lines 131-147 move to discussion

3.3 Dynamic offloading model as shown in lines 150-155

lines 157-165 move to discussion

3.4 Proposed Model Processing overview as shown in lines 299-318

lines 245 to 247 move to discussion section and discuss why you proposed your model.

lines 249 delete Use case

lines 250-256 should be moved to introduction section. Revise lines 251 remove "suffered from the proliferation by and replace with spread faster because of the strong winds. Consider revising lines 254-256 like this. The changing climate has increased the frequency of droughts resulting in huge wildfires. The poor firefighting capacities have also...........make this relevant to your study and why you are predicting with your .....algorithm.

lines 257-280 move to methods section 

lines 281-295 move to discussion section of dynamic model

lines 296-297 delete not

lines 299-318 move to results section and show one output of prediction based temperature or wind speed.

lines 320-329 move to discussion section and offer a better discussion of how your prediction model works better than the others, what is new and how it improves fire prediction than the others, try to compare your findings with other studies.

lines 334-370 move to methods section as 2.2.2.1 since they are not analyzed but just present more details about the data types

I request the authors to discuss their findings to show that the dynamic offloading model has better response to data handling processing than the others. You have already shown this in the 4 model types but you need to make a reader know why each is limited and why dynamic model is better. Lines 281-295 can offer you help you to discuss how your dynamic model functions.

Your paper is very innovative and has great information that can improve fire management in terms of prediction and prevention. Good work.

Author Response

line 18 change Edge Computing to Edge computing in the whole manuscript.

-> I changed Edge Computing to Edge computing in the whole manuscript.

line 32-34 does not make sense. I think you should insert "is not only using numerical data...." after the word analysis so that the sentence makes sense to the reader.

-> I changed the sentence. That sentence is now on line 33-35. Line 33-35 is “Nowadays, the field of analysis using not only numerical data but also image, video and audio data is expanding rapidly due to the development of the IoT and the use of numerous smart/intelligent devices.”

line 36 remove the comma after latency and do the same in the whole manuscript.

-> I removed the comma in the whole manuscript.

line 37 put the full stop after citation [3-7] and [8-13] remember to do the same in the whole manuscript

line 42 make the citations this way as required [15-17] and remember to do it in the whole manuscript.

-> I put them like [3-7] and [8-13] in the whole manuscript.

line 44 remove "However" at the start of the sentence and remove "now" after cloud computing

-> I removed “However” and “now”. That sentence is now on line 55

line 45 begin the line with "However,...........................

-> I changed sentence to the “However,….”. That sentence is now on line 56

line 48 change the sentence Other Edge computing to issues to consider are response time requirement,.........,.........,..remember to change the citations to [18,19].

-> I changed sentence. That sentence is now on line 39. Line 39 is “Issues to consider with in many papers are response time requirement, battery life constraint, bandwidth cost-saving as well as data safety and privacy dealt”

Line 50 does not make sense .......Begin with Edge computing main offloading task is dividing tasks that require.................full stop. Then begin As a result, there are offloading........

-> -> I changed sentences. That sentences are now on line 59 and line 69.

lines 53 Begin this way.... This load is a problem in areas where immediate response is needed to detect nearby objects, collect real time data and speed up complex calculations [20-22]. Then move lines 61-64 here.

-> I changed sentences to “These problems are especially important in areas where immediate response is…….”. and moved lines 61-64 to lines 68-71.

Then remove example 1 and 2 and focus on fire because your paper applies a model on fire. Just begin For example, if fire is detected...............In recent years, the seriousness of wildfires is rapidly increasing. Then continue and write one or two paragraphs explaining why the other offloading approaches are not good for forest fires or wildfires and why you propose the use of algorithm-based offloading on a risky situation that needs rapid decision action like forest fires.

line 66 remove where and replace with that

-> I removed example 1 and 2. I wrote lines 74-89 and lines 273-283 as the reason why the other offloading approaches are not good for forest fire and why we propose the use of algorithm-based offloading.

line 70 change was and replace with attempt has been made........

-> I changed to “attempt has been made”. That sentence is now on line 93.

line 72 change to Section 3 includes an .........

-> I changed to “Section 3 contains algorithms…”. That sentence is now on line 96.

line 74. Change to The conclusions of this paper are in section 5.

-> I changed to “The conclusions of this paper are in section 5.”. That sentence is now on line 98.

lines 76 should read Methods

-> I changed to method contents.

lines 77-91 should be moved to introduction to explain why other data or task offloading have failed to address forest or wildfires

-> I moved lines 77- 91 to introduction. That sentence is now on lines 41-65

line 92  should read, We propose a model.......... and delete line 93 sentence Chapter 2.1 describes edge computing based offloading in line detail

-> I changed line92 and removed line93. That sentence is now on line 103.

The methods section is not consistent.

I think lines 167-175 should be in the methods section. As the explain the edge node

I suggest that they reconsider the methods section as follows

2.1 Edge note

2.2 Master Node

2.2.1 Data centric offloading algorithm

2.2.2 Task centric offloading algorithm

2.3 Second master node

2.4 Action Node

2.5 Forest fire prediction as shown in lines 257-295

-> I changed the method section by the review presented. I composed Chapter 2 as "2. Offloading Architecture Proposed", 2.1-2.4. I moved the "2.5 Forest fire prediction" to 3.3 because this is the content of the algorithm.

3.0 Results

3.1 Offloading based on edge computing...................As shown in lines 94-103

Lines 104-121 move to discussion section

3.2 Problems of offloading based on edge computing As shown in lines 125-129

lines 131-147 move to discussion

3.3 Dynamic offloading model as shown in lines 150-155

lines 157-165 move to discussion

3.4 Proposed Model Processing overview as shown in lines 299-318

lines 245 to 247 move to discussion section and discuss why you proposed your model.

-> Chapter 3 is about the algorithm, so chapter 4 is the result and discussion section.

This paper is a model-proposed paper that was written for a modeling scenario rather than an actual experiment. In the next paper, we will add the experiment based on this content.

Therefore, this paper is composed of 4.1-4.4, and we have written discussion content together.

lines 249 delete Use case

-> deleted line 249

lines 250-256 should be moved to introduction section. Revise lines 251 remove "suffered from the proliferation by and replace with spread faster because of the strong winds. Consider revising lines 254-256 like this. The changing climate has increased the frequency of droughts resulting in huge wildfires. The poor firefighting capacities have also...........make this relevant to your study and why you are predicting with your .....algorithm.

-> I moved lines 250-256 to introduction section and removed lines 251 and revised line 254-256. That sentences are now on line 74-83

lines 257-280 move to methods section

-> I moved to 3.3 because this is the content of the algorithm.

lines 281-295 move to discussion section of dynamic model

-> Lines 281-191 moved to lines 382-392 and lines 292-295 moved to lines 228-231.

lines 296-297 delete not

-> lines 296-297 is now on line 346.

lines 299-318 move to results section and show one output of prediction based temperature or wind speed.

lines 320-329 move to discussion section and offer a better discussion of how your prediction model works better than the others, what is new and how it improves fire prediction than the others, try to compare your findings with other studies.

-> Lines299-329 moved to the 4.3 proposed model procession overview and explained in more detail with the example of temperature and wind blowing.

lines 334-370 move to methods section as 2.2.2.1 since they are not analyzed but just present more details about the data types

-> Line 334-370 moved to “3.4 Scenario of computation offloading for wild fire” because this is the content of the algorithm.

I request the authors to discuss their findings to show that the dynamic offloading model has better response to data handling processing than the others. You have already shown this in the 4 model types but you need to make a reader know why each is limited and why dynamic model is better. Lines 281-295 can offer you help you to discuss how your dynamic model functions.

-> I wrote on lines 278-282 and line 404-407. I am not sure it is you are right about this. I am not sure you are what you said on this. Please check and let me out.

Your paper is very innovative and has great information that can improve fire management in terms of prediction and prevention. Good work.

-> Thank you for your detailed review. Have a nice day.

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report

Dear Authors, I can see an effort in the review of the work, congratulations. For my part it is ready.