A Multi-Objective Roadside Unit Deployment Model for an Urban Vehicular Ad Hoc Network

Round 1

Reviewer 1 Report

- This article suggests a model for optimizing the deployment of Roadside Units (RSUs) in an urban area by applying the multi-object evolutionary algorithm. The proposed model considers three factors: the deployment cost (i.e., the number of RSUs required), RSUs coverage area, and data transmission delay. The objective is to find the best possible deployment strategy that minimizes deployment costs while maximizing coverage area and minimizing data transmission delay.

- It is worth mentioning that, the problem the authors are addressing, RSU deployment, has already been widely discussed in the literature.

- The paper is well-written and organized. The manuscript is thorough. However, there still exist some points that require to be clarified. I have listed a few comments and suggestions for the authors' consideration as follows:

1- Only reading the introduction section, I feel that the contribution of this paper is not significant once the contributions are not highlighted in this section. So, list all contributions on it.

2- It is known that NSGA-II can converge prematurely, which means that it may find suboptimal solutions far from the actual Pareto front. This is because the algorithm may get trapped in local optima or converge too quickly without exploring other regions of the search space. It would be helpful if the authors could clarify how they dealt with this issue.

3- Another issue is the stopping criterion; the NSGA-II is usually based on a predefined maximum number of generations or evaluations. However, this may not ensure convergence to the Pareto front, especially for complex problems with many objectives and decision variables. It would be helpful if the authors could clarify how they dealt with this issue.

4- Row 281: The traffic density unit is vehicles/km^2.

5- It is unclear how it was selected candidate intersections based on the NSGA-II. It would be helpful if the authors could clarify this point through an algorithm.

6- Row 408: "The trend shows the same as in the case of the simulated environment. As the number of RSUs increases, the average coverage area increases, and the average total expected delay gradually decreases." The results seem very obvious to me.

7- It is not clear what influence traffic density has on RSU deployment. It would be helpful if the authors could clarify how they dealt with this issue.

8- The study mentioned VANET technology in various parts of the text, including the title, as a potential solution for enhancing vehicle communication. However, it was not utilized in the experimental setup. It would be helpful if the authors could clarify how VANET contributes to dealing with the presented problem.

9- It would be helpful if the authors could compare the proposed solution with the following works [1], [2], [3].

10- The current state of the work and its results require further development and maturation. Therefore, it is recommended that further investigation and refinement be conducted to enhance the quality and validity of the findings.

 

[1] An evolutionary algorithm for roadside unit deployment with betweenness centrality preprocessing.

[2] Intelligent Roadside Unit Deployment in Vehicular Network.

[3] AC-RDV: a novel ant colony system for roadside units deployment in vehicular ad hoc networks.

 

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 2 Report

1. punctuation mark error

In Line 15 and line 16:’ Its deployment can effectively improve the communication performance of the network, The goal of the RSUs deployment (RSUD) problem is to install as few RSUs as possible ...............’

2. Grammar errors

 In Line170,’Firstly, take the simulation environment as an example.’ The sentence lacks subject.

3. NSGA-II algorithm is very important to solve the multi-objective optimal question. So the authors should give more detailed description about the algorithm instead of the common knowledge about it , which could convince us the truth of the study. 

Comments for author File: Comments.pdf

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 3 Report

I have only one question. Why the configuration and modeling of the vehicle-road networks did not study in the project based on the NS-2 simulator which can provide tools for the research of vehicle-road coordinated control.

Author Response

Thanks to all the reviewers for the proposed changes to this article, I have made the following changes based on the recommendations.

Point 1: I have only one question. Why the configuration and modeling of the vehicle-road networks did not study in the project based on the NS-2 simulator which can provide tools for the research of vehicle-road coordinated control.

Response 1: In this study, we use a multi-objective optimization algorithm to solve the RSUD problem, and obtain the Pareto optimal solution set. The Pareto optimal solution set contains multiple solutions with different preferences, which need to be graphically represented and selected. This is equivalent to the need for multiple configurations and modeling in the NS-2 simulator after each calculation, which consumes too much computation and time. Therefore, we choose the Matlab platform which is more conducive to algorithm optimization and result display for verification. Based on the importance of network simulator, we simulate in OMNet++ after selecting the only RSU deployment scheme for practical application.

 

Round 2

Reviewer 1 Report

Firstly, I would like to thank the authors for the revised version; it is much improved over the original submission.

Overall, I am satisfied with the revised version.