An Improved Gravitational Search Algorithm for Task Offloading in a Mobile Edge Computing Network with Task Priority
Round 1
Reviewer 1 Report
Comments and Suggestions for AuthorsThis paper proposes a multi-user mobile edge computing network model that utilises a collaborative architecture of cloud and edge servers in a mobile edge network. Multiple users have multiple subtasks to be processed, with inter-task priority constraints. The optimisation objectives are to minimise user task execution delay and server computing costs. To achieve this, we introduce an improved gravity search algorithm (IGSA) for solving the problem.
Although the work is a contribution to the area of knowledge. The following should be improved:
i) Section I includes an introduction and background. That must be separated, and the problem to be solved, how it is solved, and the contributions of the work be precise in addition to existing related works.
ii) In section II, the problem must first be formulated and modelled. Second, the system model that solves it must be presented.
iii) Section II should clarify which material is one's own creation and which material is used by other authors.
iv) Sections II and III should clarify which material is one's creation and which material is used by other authors(in other papers).
v) In section V, the authors must present results that verify that they meet the hypothesis and compare them with three results from methods in the literature, both in graphs and performance data.
Best regards
The reviewer.
Comments on the Quality of English LanguageModerate editing of English language required.
Author Response
Please see the attachment.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
Comments and Suggestions for Authors1. Summary of Contributions
This paper introduces an innovative approach to computation offloading in a three-tier mobile edge computing network. The study addresses the complex challenge of dividing user tasks into subtasks, considering both serial and parallel priority relationships. A key novelty lies in the development of an optimization model aimed at minimizing the total cost, defined by the weighted sum of system delay and energy consumption. The paper's primary contribution is the introduction of an Improved Gravitational Search Algorithm (IGSA), uniquely featuring a convergence factor for more precise force calculation and genetic algorithm crossover operations during particle generation. This approach could enhance the system's performance by optimizing task execution delays and server computing costs.
2. Weaknesses & Suggestions
(a) A thorough review of the manuscript is needed to correct typographical errors, such as "algorit0hm" found in line 322. Additionally, the text requires formatting adjustments to include proper spacing between words, as observed in lines 251 and 358. The title in Algorithm 2 should be clarified to read "Algorithm 2: IGSA". Please also rectify the presence of Chinese characters in Figure 3 for consistency and clarity.
(b) Figures 1 and 2 lack clarity and require enhancement for better visibility and understanding. On page eight, the formatting needs correction, particularly addressing the excessive blank spaces which disrupt the flow of the document.
(c) It is unclear whether the proposed algorithm maintains its superior performance when varying the number of processors and users. Additional experimental results are necessary to demonstrate the efficacy of IGSA under these varying conditions and to compare its performance with other leading algorithms in the field.
Comments on the Quality of English Language
A thorough review of the manuscript is needed to correct typographical errors, such as "algorit0hm" found in line 322. Additionally, the text requires formatting adjustments to include proper spacing between words, as observed in lines 251 and 358. The title in Algorithm 2 should be clarified to read "Algorithm 2: IGSA". Please also rectify the presence of Chinese characters in Figure 3 for consistency and clarity.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Reviewer 3 Report
Comments and Suggestions for AuthorsThe manuscript presents an improved gravitational search algorithm (IGSA) for task offloading in a mobile edge computing network with task priority. The authors have addressed an important problem in mobile edge computing by considering the task priority relationships and proposing an optimization model to minimize system delay and energy consumption. The IGSA algorithm is introduced as a solution, incorporating convergence factor and crossover operation. The simulation results indicate improved system performance compared to existing algorithms. Overall, the manuscript is well-written and provides valuable insights into the field. However, there are a few areas that require clarification and further improvement.
The introduction provides a good background on mobile edge computing and the challenges related to task offloading. However, it would be helpful to include more references to support the statements made throughout the introduction. Additionally, it would be beneficial to clearly state the research objectives and contributions of the paper at the end of the introduction section.
The manuscript briefly mentions that an optimization model is established, but the details of the model formulation are lacking. It is essential to provide a clear and comprehensive description of the optimization model, including the objective function, constraints, and variables used. This will help readers understand the problem formulation and the basis for the proposed algorithm.
The manuscript introduces the IGSA algorithm but does not provide a step-by-step description or pseudocode. It would be beneficial to include a detailed algorithmic description or pseudocode to help readers understand the implementation details of the proposed algorithm. Additionally, providing explanations or justifications for the specific modifications made to the gravitational search algorithm would enhance the clarity of the proposed approach.
The simulation results are briefly mentioned, but specific details about the experimental setup, evaluation metrics, and comparison with existing algorithms are lacking. It is crucial to provide a clear description of the experimental methodology, including the dataset used, performance metrics measured, and statistical analysis performed. Additionally, it would be beneficial to compare the proposed IGSA algorithm with state-of-the-art algorithms in the field to demonstrate its superiority.
Comments on the Quality of English LanguageThe manuscript needs extensive revision for language and grammar
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Reviewer 4 Report
Comments and Suggestions for AuthorsIn this manuscript, the authors investigate the problem of task offloading in edge computing and propose an improvement of the gravitational search algorithm, aiming to optimize the total cost in terms of delay and energy consumption.
In summary, the problem is useful and practical, and the authors explain the problem, introduce an optimization model, and provide its performance evaluations. It's a well-written, coherent work that might be useful to some readers.
However, I think that the authors could further improve it by:
a) Add an energy consumption model: the reference to energy consumption in the text (e.g., lines 8, 333, 337) makes sense to the reader that it was considered in calculating the optimization solution.
b) Clarify the role of the particle: does it represent a user devise, a task, or something else?
c) correct typos (e.g., section 2.4, replace sever with server, Figure 3: replace Chinese characters with English, etc.)
Lastly, I have a question for the authors: In practice and in real-world environments, where, when, and by whom does the proposed algorithm run?
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Reviewer 5 Report
Comments and Suggestions for Authors
In this manuscript, the authors propose a GSA-based method for computation offloading in a mobile device-edge-cloud architectural framework. They first present the mathematical modeling of each task, including the division into multiple subtasks, with serial and parallel priority among them. Moreover, the authors formulate the optimization problem of minimizing the system delay and the energy consumption, taking also into account the availability of resources and the interrelationships among the subtasks. Finally, they showcase the proposed improved gravitational search algorithm (IGSA) and its results in the task offloading problem, comparing its performance with 2 baselines. The paper is interesting and relevant to the topics of Electronics. However, in my opinion, minor revisions in this work are required:
The quality of Figure 1 and Figure 2 needs to be drastically improved. Moreover, please consider re-designing Figure 2.
Moreover, in the introduction section, the authors should additionally mention deep reinforcement learning (DRL) algorithms that are used in similar applications with joint optimization targets. For instance, except for task offloading, RL and Deep RL algorithms can be used for power control of base stations, scheduling, load balancing and many more applications in wireless networks, e.g.:
1. Chinchali, Sandeep, et al. "Cellular network traffic scheduling with deep reinforcement learning." Proceedings of the AAAI Conference on Artificial Intelligence. Vol. 32. No. 1. 2018.
2. Tang, Ming, and Vincent WS Wong. "Deep reinforcement learning for task offloading in mobile edge computing systems." IEEE Transactions on Mobile Computing 21.6 (2020): 1985-1997.
3. Liu, Qingzhi, et al. "Deep reinforcement learning for load-balancing aware network control in IoT edge systems." IEEE Transactions on Parallel and Distributed Systems 33.6 (2021): 1491-1502.
4. Spantideas, Sotirios, et al. "Intelligent Mission Critical Services over Beyond 5G Networks: Control Loop and Proactive Overload Detection." 2023 International Conference on Smart Applications, Communications and Networking (SmartNets). IEEE, 2023.
5. A. Kaloxylos, A. Gavras, D. Camps Mur, M. Ghoraishi and H. Hrasnica, “AI and ML—Enablers for Beyond 5G Networks”, Zenodo, 2020
6. Zhang, Lin, and Ying-Chang Liang. "Deep reinforcement learning for multi-agent power control in heterogeneous networks." IEEE Transactions on Wireless Communications 20.4 (2020): 2551-2564.
Comments on the Quality of English Language
There are several typos throughout the paper (missing spaces between sentences, servers instead or severs in sections 2.3 and 2.4, etc.). Please carefully proofread your manuscript.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Reviewer 6 Report
Comments and Suggestions for AuthorsThis paper aims to solve the computation offloading problem in a three-tier architecture. By using the improved GSA, this paper shows the superiority compared to previous GSA methods. There are several concerns as follows:
- Please provide the relationship between nodes and UE/Edge/Cloud in the task model.
- Could you explain the reason why the computing power of UE is ignored?
- In (21), is it correct that the normalization factor is considered only for the total cost part, not the total delay part?
- All the parameters should be clearly defined. For example, there are no definitions or there are definitions in inappropriate positions (even though readers can infer) on O_i^E, O_i^C, alpha, beta, queue_n, etc.
- Please use English words in Figure 3.
- The performance analysis is only compared with existing gravitational search algorithms. In other words, even though this paper aims to solve the computation offloading problem, there is no comparison with previous lots of computation offloading methods considering cloud, edge, and UE itself. Please provide specific reasons.
- The readability should be enhanced. Please perform the detailed proof-reading.
- in [0],Anubhav
- in scope.In [10]
- In[9]
- (GSA)and
- In[10]
- more than three times of full definition of GSA
- etc.
Comments on the Quality of English LanguagePlease perform the detailed proof-reading.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Round 2
Reviewer 1 Report
Comments and Suggestions for AuthorsThis paper proposes a new offloading algorithm called the Improved GSA-based Offloading (IGSA) algorithm. The IGSA algorithm incorporates a convergence factor to accelerate searching for optimal solutions in the search space. Additionally, it introduces crossover operations from genetic algorithms to improve the optimisation results.
This new work version presents significant improvements, although there are still things to debug. Such as the following:
i) Review the connection between paragraphs in all sections. For example, page 2, line 83. Page 7, lines 220-222. And check the "-" symbol on page 4 (line 137).
ii) Improve section 5 (Results) to support the verification of the solution to the problem and the improvement of the proposed algorithm versus others already used.
iii) Add a results discussion section to support the conclusions section and the results obtained.
Comments on the Quality of English Language
Moderate editing of English language required
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Reviewer 2 Report
Comments and Suggestions for AuthorsThis revised version is well-written, offering detailed insights into the design and effectively addressing my concerns.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Reviewer 6 Report
Comments and Suggestions for AuthorsMy previous comments are well-considered in the revised paper.
I have no additional comment.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Round 3
Reviewer 1 Report
Comments and Suggestions for AuthorsThis paper investigates the problem of computation offloading in a three-tier mobile edge computing network composed of multiple users, multiple edge servers, and a cloud server. In this network, each user’s task can be divided into multiple subtasks, with serial and parallel priority relationships existing among these subtasks. An optimization model is established with the objective of minimizing total user delay and processor cost, under constraints such as the available resources of users and servers and the interrelationships among the subtasks. This paper proposes a new offloading algorithm called Improved GSA-based Offloading (IGSA) algorithm.
The document presents improvements compared to the current version, although it also has deficiencies to correct. Such as:
i) the phrase on line 53 is redundant.
ii) Present the objective in a better way, do not include this in the contributions.
iii) In section 3, System model, include reference to formulas that are not your own.
iv) Section 5 should be renamed to: Simulations and Results.
v) Add Discussion section, and Improve Conclusions section.
Best Regard
The reviewer
Comments on the Quality of English Language
Moderate editing
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
