Multi-Objective Service Placement Scheme Based on Fuzzy-AHP System for Distributed Cloud Computing

Round 1

Reviewer 1 Report

1)There are two research problems that you want to solve: 1. metrics vs objectives; and 2. balance conflicting objectives. I think these two problems are not explained clearly. Why are they so important? No previous studies have been done before?

 2) You did not provide sufficient literature review in your "Related Works". For example, you described service placement in 2.1 with only few studies. Why you used MCDM is still an unknown issue. I don't see it in your 2.2. 

3)Generally, Fuzzy numbers are assigned for each decision. I don't see them in your paragraph. There is only a very unclear figure.

4) What is the ranking of metrics in your experiment? In AHP, it is important to form your decision alternatives, I only see A1- A5. In addition, you only measure the proposed method with other methods based on execution time and energy efficiency. You did not answer the research questions.

5) In your Table 10, Scenario 4 and 7 are same settings. 

6) There are tables (table 1 to table 5) with few lines of explanations.

7) A lot if grammar issues and typos have to be addressed. 

8) Make sure that abbreviation has the full name. For example, what is VM?

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

The research topic in this paper is very important. Also, the authors had a very good background introduction. For example, Table 1-3 give a good and clear summary of the related work. In addition, the authors clearly illustrated their methodology, evaluation, and the result analysis. I only have some minor comments for improving this paper for the publication.

The title of this paper shows this work focuses on the edge-enabled distribution cloud computing. However, it seems to me that nothing is special for the edge computing. This work just gives a general resource allocation and job dispatching approach. Thus, it is better to revise the title.

The resolution of all figures in this paper is pretty bad. Are they from reference or made by the authors? If they are from the reference, the authors need to give the reference. If made by the authors, high resolution figures are expected.

A better explaination of fuzzy system is expected.

Is the execution time used in the proposed methodology the average time or the worst-case execution time (wcet)? Usually, WCET is used in this kind of research.

Why did the authors choose the evaluation environment in Table 9?

How did the authors collect these performance data? any tool?

The rules in Table 10 are either "very high" or "very low". Why not "medium"? Does fuzzy system only allows "very high" and "very low"?

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

There is a huge improvement in the revised manuscript, but I still have some problems to ask.

You have mentioned RU in the related work section, and you described the factors in Table 7. However, from your AHP structure (Figure 6), I don't see that you use RU. Then, in line 866, you wrote: "In equation for RU, we decided to two specific metrics, CU and MU for RU in this paper." It's very confusing. I believe that you need to explain more because your M1 is resource utilization of VM, which seems similar to me. After you deleted edge computing of the title, you mentioned edge environment or edge computing a few times in the paragraphs. I think these are now unnecessary words.  The figures are still having bad resolutions.

Author Response

Please see the attachment

Author Response File: Author Response.pdf