Next Article in Journal
Out-of-Plane Behavior of Masonry Prisms Retrofitted with Shape Memory Alloy Stripes: Numerical and Parametric Analysis
Next Article in Special Issue
Multi-Feature Transformer-Based Learning for Continuous Human Motion Recognition with High Similarity Using mmWave FMCW Radar
Previous Article in Journal
Applying Artificial Intelligence to Wearable Sensor Data to Diagnose and Predict Cardiovascular Disease: A Review
Previous Article in Special Issue
An Infrastructure for Enabling Dynamic Fault Tolerance in Highly-Reliable Adaptive Distributed Embedded Systems Based on Switched Ethernet
 
 
Article
Peer-Review Record

Reinforcement Learning-Aided Edge Intelligence Framework for Delay-Sensitive Industrial Applications

Sensors 2022, 22(20), 8001; https://doi.org/10.3390/s22208001
by Muhammad Zubair Islam 1, Shahzad 1, Rashid Ali 2, Amir Haider 1,* and Hyung Seok Kim 1,*
Reviewer 1: Anonymous
Reviewer 2:
Sensors 2022, 22(20), 8001; https://doi.org/10.3390/s22208001
Submission received: 15 September 2022 / Revised: 13 October 2022 / Accepted: 17 October 2022 / Published: 20 October 2022
(This article belongs to the Special Issue Feature Papers in the Sensor Networks Section 2022)

Round 1

Reviewer 1 Report

- Background and related work about software-defined networking (SDN) are missing (in the introduction and related work). SDN is mentioned in Fig. 3, but it is not explained. How is it related to Tactile Internet?

- Please explain why the algorithms or models (LSTM, Q-learning) were selected. There are more advanced algorithms and models for predictions and learning, e.g., Double Q-learning, Deep Deterministic Policy Gradient.

- The experiment is missing the comparison with other methods or algorithms. This is important to show the advantages of the used algorithms or models.

- Some Abbreviations are used without full explanation in the text or in the Abbreviations table, e.g., in Table 1, SNR, SIG, etc. 

- Some content is similar or exact to the published publications (checked by Turnitin. Without the references, the similar rate is 25% - which is high).

- References: The format is not consistent and the references are missing some information, e.g., volume. Please check all references. For example, [29] and [30] have the same authors but their names are different. [30] is missing the volume. [2] Ieee?

- Presentation:

+ Table 1, In the Effectiveness column, Sig -> SIG, Not Sig -> NOT SIG, mixed -> Mixed.

+ After an equation, the "where" should not have a tab space. After an equation, some have a comma, some have not.

+ In Figures, the text is small, which is not easy to read. The Figures are also small in the experiment section.

+ The hyperlink is broken, e.g., (http://mininet.org/,accessedon15September2021), in which accessedon15September2021 has no space between the words, and when clicking the link, it cannot open.

+ Please revise the writing also. For example, after a dot, it has to have a space (see the caption of Table 1).

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

- The abstract can include numerical results obtained.

- There have been many related works. What are the gaps identified that motivated the current research?

- Some of the recent works such as the following can be discussed A survey on Zero touch network and Service (ZSM) Management for 5G and beyond networks, AI and 6G Security: Opportunities and Challenges, 

- Proposed framework section should include a discussion on the novelty of the proposed RNN based approach.

- COmpare the results obtained with recent state of the art.

- What is the computational and communication cost of the proposed approach?

- What are the threats to validity of the proposed approach?

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

The authors have addressed my comments thoroughly.

Reviewer 2 Report

The authors have addressed all the comments. I have no further comments/queries.

Back to TopTop