Joint Optimization of Service Migration and Resource Allocation in Mobile Edge–Cloud Computing

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The submitted paper presents a valuable effort in the field of edge network resource management. The literature review covers most of the present works in the field and targeted gaps. The writing style is clear and concise, making the complex details easy to understand. The problem formulation and simulation procedure are clearly described, and the results are well presented.

There is just a minor issue in line 315 where a sentence (‘Vtk+1 will discuss later’) seems to be incomplete.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

The paper describe an advanced framework, optimized through a Deep Reinforcement Learning strategy and grounded in a Markov Decision Process that dynamically adjusts service migration and resource allocation strategies. This enhanced approach enables continuous system monitoring, proficient decision-making, and iterative policy refinement, significantly improving operational efficiency and reducing response times in Multi-access Edge Computing  environments. The study deal with the practical deployment of edge computing technologies providing contributions to both theoretical insights and practical applications.

I suggest editor to accept the manuscript since it is well written and well structured.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

This paper explores the dynamic migration and resource allocation problem in mobile edge computing systems. The authors propose an Advantage Actor-Critic framework-based method to address the challenges from dynamic computational demands and user mobility. The authors evaluated their system modeling and optimization method through simulation. Overall, I found this paper timely and significant to edge computing research. The authors did a fine job proposing a system model and a framework to solve the optimization problem.

Here are a few comments to help improve the paper - 

1. The authors should elaborate and justify the sensibleness of their assumed system model. The authors assume a centralized entity that has all the knowledge of the users and the tasks they run. Is such a system modeling realistic in practical deployment? What's the overhead of gathering all those information into a centralized place? What's the privacy and security implications of such centralized decision making?

2. In the authors' evaluation, the numbers of edge servers, users, and data sizes are too small. For instance, a typical 5G RAN cell tower serves on the orders of thousands if not more users. The authors need to increase the size of their variables in the simulation by multiple order of magnitude.

3. In the authors' evaluation, one thing worth adding is the cost of proposed migration and allocation algorithm. How long does the solving the optimization problem itself take? How much compute resources does the allocation algorithm need? How does this overhead impact user experience?

Comments on the Quality of English Language

The quality of English is fine. Minor proofreading needed.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 3 Report

Comments and Suggestions for Authors

The authors have effectively addressed the concerns raised in my previous review.