Next Article in Journal
Forecasting the Unseen: Enhancing Tsunami Occurrence Predictions with Machine-Learning-Driven Analytics
Next Article in Special Issue
Improving Real-Time Economic Decisions Through Edge Computing: Implications for Financial Contagion Risk Management
Previous Article in Journal
Combining the Strengths of LLMs and Persuasive Technology to Combat Cyberhate
Previous Article in Special Issue
Fine-Tuning Network Slicing in 5G: Unveiling Mathematical Equations for Precision Classification
 
 
Article
Peer-Review Record

Investigation of Multiple Hybrid Deep Learning Models for Accurate and Optimized Network Slicing

Computers 2025, 14(5), 174; https://doi.org/10.3390/computers14050174
by Ahmed Raoof Nasser 1 and Omar Younis Alani 2,*
Reviewer 1: Anonymous
Reviewer 2:
Computers 2025, 14(5), 174; https://doi.org/10.3390/computers14050174
Submission received: 24 March 2025 / Revised: 17 April 2025 / Accepted: 26 April 2025 / Published: 2 May 2025

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

While the study presents interesting ideas and promising results, significant improvements in clarity, methodology presentation, experimental analysis, and writing quality are needed before the paper can be considered for publication.

Relevance and Contribution:

-The paper tackles an important problem in 5G and beyond networks by investigating hybrid deep learning models for network slicing.

-The integration of CNN with LSTM, RNN, and GRU alongside Bayesian optimization for hyperparameter tuning is a promising approach.

Clarity and Structure:

-The manuscript is quite dense and technical; consider reorganizing sections to enhance clarity and ensure that the reader can easily follow the research flow.

-The introduction should more clearly articulate the research gap, contributions, and the motivation behind choosing specific hybrid models.

Methodological Details:

-The explanation of hybrid model architectures and the hyperparameter tuning process is overly detailed and sometimes redundant. A more concise description, possibly with a flowchart or schematic summary, would help.

-It is not fully clear how the datasets were preprocessed or why they were specifically chosen. More details on these aspects would strengthen the methodology.

Experimental Evaluation:

-Although extensive experimental results are provided, the paper lacks a thorough discussion of the computational complexity, training time, and scalability of the proposed models.

-The evaluation could benefit from additional analysis of statistical significance and error margins to better substantiate the performance claims.

-The comparison with state-of-the-art methods should be expanded to include more detailed baseline descriptions and a discussion on why the proposed models outperform these methods.

Figures and Tables:

-Some figures and tables (e.g., performance metrics, network simulation results) are not fully self-explanatory. Improve labelling, captions, and integration with the text to ensure they effectively support the discussion.

-Ensure consistency in the presentation of equations and graphical elements throughout the manuscript.

Writing and Language:

-The manuscript would benefit from a thorough language and style revision to address grammatical errors and improve overall readability.

-Streamline the literature review to focus on the most relevant works and clearly define how the current work fills the existing research gap.

Practical Implications and Limitations:

-A discussion on the practical deployment challenges, potential limitations, and future research directions is missing. Including these elements would provide a more balanced perspective on the contribution of the work.

 

 

Author Response

The reply is in the attached file

Author Response File: Author Response.docx

Reviewer 2 Report

Comments and Suggestions for Authors

The paper presents a  hybrid deep learning model combining CNN, LSTM, RNN, and GRU. This combination has strong potential for improving network slicing in 5G, offering a fresh perspective on using these technologies together for accurate network slice classification. Contributions of the paper are clearly summarized, offering a concise and structured overview of the work.  The reported results, especially the 99.31% accuracy achieved with the CNN-GRU hybrid model, highlight the effectiveness of the proposed approach. A more in-depth comparison with existing state-of-the-art network slicing models would help emphasize the novelty and advantages of your proposed hybrid approach.  A brief section on potential future improvements or extensions of your hybrid model, such as the integration of other deep learning methods or further optimizations, could provide additional insight into how this research can evolve.

Comments on the Quality of English Language

Good

Author Response

Please see the reply in the attached file 

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

All comments raised in round 1 were addressed- recommend acceptance 

Back to TopTop