Enhancing Indoor Localization Accuracy through Multiple Access Point Deployment
Round 1
Reviewer 1 Report
Comments and Suggestions for AuthorsThis study significantly enhances indoor localization accuracy by deploying multiple Access Points (APs). The core of the research focuses on improving indoor localization using a Radio Environment Map (REM) and machine learning algorithms. By integrating RSSI measurements from multiple APs, localization error rates were reduced, and positional accuracy was improved. In my opinion, this work falls within the Electronics and should be published with some modifications. The following are major suggestions and explanations that must be addressed before publication.
Comments:
1. The content mentioned in the first sentence of the third paragraph in the introduction section lacks appropriate references. It is recommended to add the relevant references.
2. At the end of the introduction, the authors state that “thoroughly evaluate several Machine learning algorithms, including the random forest regressor, Extra-Trees Regression (ETR), an artificial neural network (ANN), and AdaBoost regression.,” but the “AdaBoost regression” is not covered in the evaluation section. Please check and revise the relevant part.
3. In the section describing the integration of the radio environment and enhanced indoor positioning system, it is suggested that the author adds a detailed system block diagram. This will help readers to better understand the structure and operation principle of the system.
4. The current introduction of Figure 2 is relatively brief. It is recommended to provide a more detailed description of the “Proposed Methodology for REM and Indoor Localization”. In addition, the corresponding section recommends following the framework diagram in Figure 2 to ensure that the reader can clearly understand the whole process of the method.
5. In Section 5, it is suggested to explain in detail how the author utilizes these regression techniques to refine the two-dimensional plane map of the radio environment mapping (REM). The focus should be on introducing the specific application methods and steps, rather than merely explaining their concepts and advantages.
6. The author mentions using ensemble learning methods, specifically Extreme Trees Regression (ETR) and Random Forest, in Section 4, but does not reflect the use of Random Forest in the specific description. It is recommended to supplement the relevant content and describe in detail the application of the Random Forest method.
7. In the numerical analysis section, the author evaluated regression techniques such as ETR, Random Forest Regression, and Artificial Neural Network. However, in Section 5, AdaBoost Regressor, Support Vector Regressor, and Bagging Regressor were also mentioned. Why did the author choose to evaluate only three methods? What are the reasons for introducing the other techniques without evaluation? It is suggested to provide the specific reasons for not evaluating the other methods.
8. The description of Table 2 in the numerical analysis section is currently relatively brief. It is recommended to add more detailed descriptions, including specific numerical results, analysis processes, and conclusions, to help readers better understand the research results.
9. The author mentions that the ETR model is optimized through 10-fold cross-validation and hyperparameter tuning but does not provide a detailed introduction. It is recommended to supplement the relevant content and explain in detail the specific methods and steps of cross-validation and hyperparameter tuning.
10. Additionally, please note the following writing details:
(1) For abbreviations appearing for the first time, such as RMSE (Root Mean Square Error), provide detailed explanations;
(2) Adjust the scale of Figure 5,
(3) The font size in Figure 8 is too small, it is recommended to enlarge the font size to improve readability;
(4) Improve the language level of the article to ensure that the wording is professional, precise, and clearly articulated.
11. It is recommended that the authors include literature highly relevant to artificial neural network. These references are closely related to the topic of this paper and can provide readers with a more comprehensive background and understanding:[1, 2]
[1] Gao, X. X. Ma, Q. Gu, Z. Cui, W. Y. Liu, C. Zhang, J. J. Cui, T. J., "Programmable surface plasmonic neural networks for microwave detection and processing," Nature Electronics, vol. 6, no. 4, pp. 319-328, Apr 2023, doi: 10.1038/s41928-023-00951-x.
[2] C. Liu et al., "A programmable diffractive deep neural network based on a digital-coding metasurface array," Nature Electronics, vol. 5, pp. 113–122, 2022, doi: 10.1038/s41928-022-00719-9.
Comments for author File: 
Comments.pdf
Author Response
Please go through the docx file of the review response.
Author Response File: 
Author Response.docx
Reviewer 2 Report
Comments and Suggestions for AuthorsIn this article, the authors investigate how integrating numerous access points can improve indoor localization accuracy. This method is based on the theory that using RSSI measurements from a wide range of access points can dramatically lower localization error rates and enhances position accuracy. In order to verify this theory, they build an REM by combining RSSI information from many access points gathered by a mobile robot in the intended interior setting. We evaluate several machine learning regression techniques using an improved dataset. The author's work is not bad but still, there are several limitations that must be removed, such as:
1. The abstract should concisely outline the primary contributions and discoveries of the research. Presently, the text is too wordy and would benefit from a greater emphasis on the primary findings and consequences of the study.
2. The related works section should offer a more exhaustive evaluation of the current body of literature. Identify the deficiencies in prior research that your work focuses on and explain how your methodology distinguishes itself from or enhances existing methodologies. The authors can also consider the work in “Mobile robot localization: current challenges and future prospective”, and “Extended Kalman filter-based localization algorithm by edge computing in wireless sensor networks”.
3. The authors should provide a more comprehensive elucidation of the approach employed for the collection and processing of data. Provide detailed information regarding the experimental configuration, methods used to collect data, and any underlying assumptions made during the study.
4. The explanations of the machine learning techniques, namely the Extra-Trees Regressor (ETR), ought to be more comprehensive.
5. Elucidate the procedure of dataset construction utilizing the K-nearest neighbor search. Enhancing transparency in the process of collecting data points and building samples could be achieved by providing pseudocode or flowcharts to facilitate comprehension.
6. The authors should provide a rationale for selecting the performance criteria, such as RMSE and Euclidean distance, that were used to evaluate the models. Explain the suitability of these indicators and how they offer a significant evaluation of the model's performance.
7. The results section should have a more comprehensive analysis of the discoveries. Examine the consequences of the reported error rates and their influence on the practical implementation of the suggested localization system.
8. Further enhance the precision and comprehensibility of figures and graphs. Make sure that all visual components are easily legible, featuring well-defined axes, labels, and captions that offer ample context for the shown data.
9. Perform an in-depth comparative investigation of your suggested methodology and other pre-existing methodologies. Provide an analysis of the advantages and disadvantages of each strategy, and explain how your way offers enhancements.
10. Moreover, elaborate on the future work section to delineate possible expansions of your research. Examine the potential application of the suggested methodology to different indoor settings, possible enhancements in data gathering methods, and alternative machine learning models that could be investigated.
11. There are several grammatical mistakes that must be corrected with detailed proofreading.
Comments on the Quality of English Language
Extensive editing of English language required
Author Response
Please go through the docx file of the review response.
Author Response File: Author Response.pdf
Round 2
Reviewer 2 Report
Comments and Suggestions for AuthorsNo further comments from my side.
