Review Reports

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The study presents an innovative application of explainable multi-agent reinforcement learning (MARL) for spectrum situational awareness, addressing key limitations of conventional spectrum monitoring techniques. Below are some suggestions and questions.

1. The study could benefit from a more comprehensive comparison with existing methods (e.g., other reinforcement learning-based or heuristic-based spectrum monitoring techniques).
2. Have the authors tested SmartScan across different environments?
3. While real-time performance is highlighted, some analysis of computational overhead (e.g., inference latency per decision step) would be helpful.
4. Providing numerical performance comparisons with baseline methods would further strengthen the argument for SmartScan's superiority.

Author Response

Please see the attachment

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

This topic is very interesting. Some questions and suggestions are as follows.

(1)What is fundamental of SmartScan framework? It should be explained in the paper.

(2)What is the function of FFT Frame?

(3)Algorithm 1 only contains word descriptions. It lacks math model for support. It is not like a algorithm, but a procedure.

(4)The details of DQN should be added for better understanding.

(5)In equation（1）, the math symbols * in the Q* lacks definitions for illustration.

 

Author Response

Please see attachment.

Author Response File: Author Response.pdf