Next Article in Journal
GTrXL-SAC-Based Path Planning and Obstacle-Aware Control Decision-Making for UAV Autonomous Control
Previous Article in Journal
Safe and Optimal Motion Planning for Autonomous Underwater Vehicles: A Robust Model Predictive Control Framework Integrating Fast Marching Time Objectives and Adaptive Control Barrier Functions
 
 
Article
Peer-Review Record

Joint Optimization of Task Completion Time and Energy Consumption in UAV-Enabled Mobile Edge Computing

by Hanwen Zhang, Tao Chen *, Bangbang Ren, Ruozhe Li and Hao Yuan
Reviewer 1: Anonymous
Reviewer 2: Anonymous
Reviewer 3:
Submission received: 24 February 2025 / Revised: 17 March 2025 / Accepted: 1 April 2025 / Published: 3 April 2025

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The topic has no value. There have been many articles on problem research published 4 to 5 years ago.

Comments on the Quality of English Language

common

Author Response

请参阅附件。

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

This paper proposes a Lyapunov optimization and reinforcement learning (RL)-based framework (LyraRD) for the joint optimization of task completion time and energy consumption in UAV-enabled mobile edge computing (MEC). The authors introduce a maximum completion time (makespan) metric to quantify task latency and use Lyapunov optimization to ensure energy queue stability. The problem is decomposed into per-slot deterministic optimization subproblems, solved via a novel actor-critic RL architecture. Simulation demonstrates improvements over benchmarks in balancing energy consumption and latency. This paper is well-written and novel. Here are some comments to further improve. 

1. While the integration of Lyapunov optimization with RL is good, the manuscript insufficiently differentiates its approach from prior works. The authors should better clarify how their model explicitly addresses sequential task execution delays and non-uniform energy balancing compared to existing solutions. 

2. Adding recent advances in UAV communication optimization and RL could strengthen the literature review.
[1] Securing UAV-Aided NOMA Wireless Powered Communications via Artificial Noise (https://ieeexplore.ieee.org/abstract/document/10908552)
[2]  Collaborative Ground-Space Communications via Evolutionary Multi-objective Deep Reinforcement Learning
[3] Evolutionary reinforcement learning with action sequence search for imperfect information games

3. The LyraRD algorithm’s design lacks some implementation specifics. For example, what are the input features (e.g., UAV energy states, task queues) to the DNN? How is the action space quantized (lines 5–18, Algorithm 2)? What hyperparameters (learning rate, reward scaling) and network architectures (layers, activation functions) are used?

4. The evaluation assumes fixed UAV trajectories, which seems to oversimplify real-world mobility dynamics. Could you justify this?

5. The experiments use a static UE distribution and simplistic mobility models, limiting applicability to dynamic IoT environments (e.g., vehicular networks). Extensions to heterogeneous UAV capabilities or multi-agent coordination are not discussed. These points should be clarified and added in future work. 

6. It would be better to solve some typos and grammar issues. 

Comments on the Quality of English Language

Please fix some typos, such as 

1.1 Real-time task arrival ->1.1 Real-time Task Arrival 

Existing work focuses on the problem-> Existing works focus on

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

The authors have presented a study of a UAV-enabled MEC system that satisfies long-term energy consumption constraints with high response efficiency. I have the following questions:

1- How many UAVs do you need to cover the simulation area that you have simulated during 24 hours a day?

2- What are the frequency bands where you can have available 180MHz?

3- Why have you chosen height=100m?

4- There are a few typos. For istance, in pag. 3 modle -» model. Appendix H f11 is repeated

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

This article is not worth publishing. Because this topic has been studied a lot several years ago, and the research results are very rich and effective. It is not to add some cutting-edge literature to supplement the cutting-edge and innovative nature of your research.

Comments on the Quality of English Language

OK.

Reviewer 2 Report

Comments and Suggestions for Authors

All my concerns have been addressed, thanks to the authors' efforts.

Back to TopTop