Adaptive Resource Allocation for Emergency Communications with Unmanned Aerial Vehicle-Assisted Free Space Optical/Radio Frequency Relay System

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

This paper is well written, which proposes a UAV-coordinated K-Means MADDPG (KMADDPG) to maximize the number of completed tasks while 

prioritizing high-priority tasks to solve the scarcity of communication resources in disaster-affected areas.

The main concerns are listed as follows.

1.There are several MADRL algorithms out there, so why choose MADDPG?

2. The disaster  area is set to be 200m×200m. The authors should explain the reason.

3. In the air-to-air channel, the turbulence effect is ignored. The FSO channel should be modeled as LN, GG Malaga turbulence models, where there are always fading events. This paper only considers the main effect of attenuation.

4. For DRL, the reviewer suggests elaborating the training process and testing process separately.  offline training is preferred, and how to design the training dataset and loss function to guarantee the generalization capability is the key issue. 

5. Is the training process and testing process sharing the same sets?

6. In this model, does the action of each agent have an impact on the other agent's state? 

7. Thera are also some relavent REFs, on UAV and MADDPG, the reviewer suggests to add them to make the artcle more readable.

[1]Y. Yang, T. Song, J. Yang, H. Xu and S. Xing, "Joint Energy and AoI Optimization in UAV-Assisted MEC-WET Systems," in IEEE Sensors Journal, vol. 24, no. 9, pp. 15110-15124, 1 May1, 2024, doi: 10.1109/JSEN.2024.3378844. 

[2]S. Guo and X. Zhao, "Multi-Agent Deep Reinforcement Learning Based Transmission Latency Minimization for Delay-Sensitive Cognitive Satellite-UAV Networks," in IEEE Transactions on Communications, vol. 71, no. 1, pp. 131-144, Jan. 2023.

[3]Z. Qin, Z. Liu, G. Han, C. Lin, L. Guo and L. Xie, "Distributed UAV-BSs Trajectory Optimization for User-Level Fair Communication Service With Multi-Agent Deep Reinforcement Learning," in IEEE Transactions on Vehicular Technology, vol. 70, no. 12, pp. 12290-12301, Dec. 2021.

 

Author Response

Dear Reviewer, 

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

The authors explore the issues of edge offloading and resource allocation in UAV-supported emergency communication systems. Based on the K-means algorithm and the Multi-Agent Deep Deterministic Policy Gradient (MADDPG) algorithm, the authors propose a UAV-coordinated K-Means MADDPG. The reviewer has the following questions regarding the research described in the paper.

1. It is recommended to place Figures 3 and 4 after the first paragraph of subsection 5.2. The same applies to Figures 5 and 6. It is suggested to place the images within the corresponding subsections.

2. In subsection 5.2, Figure 3 illustrates the trend of average system rewards, while Figure 4 shows the trend of success rates. Please ensure accurate expression in the text.

3. In subsection 5.4, did the authors optimize the ratio of UAVs to MUs in the proposed algorithm? It is suggested that the authors provide more detailed explanations and clarifications in the text.

Comments on the Quality of English Language

Moderate editing of English language required.

Author Response

Dear Reviewer,

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

The revised paper looks good to me. The reviewer has no more comments.