Path Load Adaptive Migration for Routing and Bandwidth Allocation in Mobile-Aware Service Function Chain
Round 1
Reviewer 1 Report
The authors propose a migration strategy for user mobility in networks, with a PathLoad Adaptive Routing and Bandwidth Allocation (PLARBA) algorithm with significant advantages in reducing the user’s waiting time, decreasing migration failure rate, and improving QoS. The results, of the experiments reported in the paper, performed in simulation were satisfactory. The article presents an interesting option in the field of action, it is pertinent and in general it is a good document, however it has some details to be improved in terms of presentation and organization of the information, which are presented below.
1. In the abstract the authors mention "The experimental results show that the proposed algorithm has significant advantages in reducing the user’s waiting time, decreasing migration failure rate, and improving QoS compared to the soft migration strategy and two RBA algorithms ", I consider It is convenient to quantify which was the reduction of the user's waiting time, and the other improvements, so that it is not something subjective, because as it is it would be an opinion of the authors.
2. Between lines 49-68 some related works are presented, however, there is no conclusion regarding what they found about. I think some text summarizing what was found is convenient.
3. In line 89 the authors mention "There is no research on migration in this scenario", I think it is convenient to clarify that they did not find in the review they performed.
4. In the “related work” section, it is important that in each of the subsections (2.1 and 2.2) explain what are the gaps of the consulted works, detailing which are the limitations or inconveniences with the current proposals.
5. Please, correct in figure 1 the word "Vitrual.
6. It is recommended to improve Figure 6, so that the text in some components of the flow chart is not left out (of each component)
7. I believe that point 4.2, which represents the major contribution of the paper, should be improved in terms of organization and numbering. There is a numbering from (1) to (4) that would correspond to the parts of the migration algorithm. Then on line 344 there is another similar numbering (1) and (2), to describe the PLARBA algorithm.
8. Please improve the overlay text on lines 456 and 457.
9. In figure 9, I recommend enlarging the size of each square, perhaps presenting horizontally a maximum of 2 figures, leaving 1 of the 3 below.
Author Response
Please see the attachment.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
The following changes/clarifications are needed:
-the organization of the manuscript should be improved; in particular some contributions of the introduction should be inserted in the related work section;
-An example illustrating the proposed solution should be introduced before introducing the system model;
-The study assumptions should be clearly expressed and discussed; for instance is not clear if the VNFI can be shared by more than one SFC [1];
-the computational complexity of the proposed heuristic should be discussed;
-the choice of the benchmark solutions should be justified;
-English and writing style should be improved.
Minor Change
The formulas should be better written as well as the resolution of the figures should be improved.
[1] V. Eramo, F.G. Lavacca: Proposal and Investigation of a Reconfiguration Cost Aware Policy for Resource Allocation in Multi-Provider NFV Infrastructures Interconnected by Elastic Optical Networks, IEEE Journal of Lightwave. Technology, vol. 37, no. 16, August 2019, pp. 4098-4114
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Round 2
Reviewer 1 Report
The authors have considerably improved the article, correcting what was requested.Author Response
Thank you for your review very much. Thank you for your nice suggestions.
Reviewer 2 Report
Thanks for addressing all of my comments. A big effort has been done. However my comment about the assumptions should be better addressed.
In particular I report the following suggestions:
-you should insert all of the assumptions in a section; for instance you could inset them at the beginning of the Section 3;
-you should highlight the limits of the assumptions and report references for them;
-the assumptions not to share the VNFIs may lead to an inefficiency in using the cloud resources as highlight in [1] when traffic changes occur; for the support of some common user service, you could instantiate a same VNFI and when the user is moving the reconfiguration could occur by re-routing the SFC (with lossless mechanism) only without migrating VNFIs. You should discuss this aspect and better motivate the choice not to share VNFIs.
[1] V. Eramo, F.G. Lavacca: Proposal and Investigation of a Reconfiguration Cost Aware Policy for Resource Allocation in Multi-Provider NFV Infrastructures Interconnected by Elastic Optical Networks, IEEE Journal of Lightwave. Technology, vol. 37, no. 16, August 2019, pp. 4098-4114
Author Response
Please see the attachment
Author Response File: Author Response.pdf
Round 3
Reviewer 2 Report
Thanks for addressing my last comments and for your patience.
In the final version you could insert a reference on the sharing of VNFI (lines 166). Reference [1] is more than appropriate because a cloud resource allocation solution with shared VNFIs is proposed in an NFV network in which the NFVI-PoPs are interconnected by an elastic optical network.
I you do not like this reference, you can insert another reference in which solutions with shared VNFIs are proposed.
[1] V. Eramo, F.G. Lavacca: Proposal and Investigation of a Reconfiguration Cost Aware Policy for Resource Allocation in Multi-Provider NFV Infrastructures Interconnected by Elastic Optical Networks, IEEE Journal of Lightwave. Technology, vol. 37, no. 16, August 2019, pp. 4098-4114
