A Pseudo-Satellite Fingerprint Localization Method Based on Discriminative Deep Belief Networks

Round 1

Reviewer 1 Report

Your research focuses on pseudo-satellite indoor localization technology and explores the issue of pseudo-satellite indoor localization in non-line-of-sight environments, which is undoubtedly of great theoretical and practical value. Regarding the content of the thesis, it is suggested to further improve in the following aspects:

Point 1:

In analyzing the results of the Pseudo-satellite C/N0 time stability test, the analysis is not comprehensive enough to assess the characteristics of the data distribution using statistics like the maximum value, minimum value, and standard deviation.

Point 2:

4.1 The Test Environment and Test Setup section provides an insufficient description of the test environment, lacking details on the general characteristics and specific parameters (e.g., size of the test area).

Point 3:

In the performance evaluation of static positioning accuracy, the accuracy of the selected reference datum and the rationale for selecting the test points are not described in detail.

Your research focuses on pseudo-satellite indoor localization technology and explores the issue of pseudo-satellite indoor localization in non-line-of-sight environments, which is undoubtedly of great theoretical and practical value. Regarding the content of the thesis, it is suggested to further improve in the following aspects:

Point 1:

In analyzing the results of the Pseudo-satellite C/N0 time stability test, the analysis is not comprehensive enough to assess the characteristics of the data distribution using statistics like the maximum value, minimum value, and standard deviation.

Point 2:

4.1 The Test Environment and Test Setup section provides an insufficient description of the test environment, lacking details on the general characteristics and specific parameters (e.g., size of the test area).

Point 3:

In the performance evaluation of static positioning accuracy, the accuracy of the selected reference datum and the rationale for selecting the test points are not described in detail.

Author Response

请参阅附件。

Author Response File: Author Response.docx

Reviewer 2 Report

Comments for author File: Comments.pdf

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 3 Report

In this paper, the indoor array pseudo-satellite C/N0 signal strength is utilized for fingerprint matching localization in a non-line-of-sight environment, based on the principle of discriminative depth belief networks. Compared with other methods, the experimental results have been enhanced and hold reference value.

Question 1: The title of the article and the description of the discriminative depth belief networks in the body of the article need to be consistent, and the authors are advised to proofread further.

Question 2: Ensure that keyword abbreviations introduced for the first time in the text are clearly defined. For example, DDBN and NLOS in line 34.

Question 3: Although you have already cited some important indoor positioning technology research results in lines 48-51 of the introduction, it is recommended to include references related to 5G mobile communication technology as one of the significant indoor positioning technology methods in this section.

Question 4: The text labeling in Figure 3 of the paper is presented in a text box. Please consider adjusting the text box background to a transparent style. This adjustment will enable the text to blend more seamlessly with the graphic background.

Further improvement is needed in language presentation and literature citation.

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 4 Report

In this paper, an innovative discriminative depth belief network method is proposed to achieve high-precision localization using pseudo-satellites in non-line-of-sight environments. Regarding the content of the paper, here are the suggestions that may help to further enhance the quality:

1. In lines 18 and 19, the consecutive use of the word "therefore" in two places to lead to a conclusion or inference is too redundant. It is suggested that an attempt be made to reorganize the sentence structure to ensure that the use of logical connectives is more concise and appropriate.

2. In line 217, the paper describes the implementation and effectiveness of the algorithm in each application area. To enhance the academic rigor and scientific quality of the paper, I would suggest adding appropriate literature citations to each application area section.

3. In line 362, the details of the presentation of the results need to be corrected. The paper states that "the results of the experiment are shown in Figure 9"; however, based on the context and the actual layout of the graphical content in the paper, the results should be presented in a series of graphs from Figure 9 to Figure 12. Therefore, it is suggested that you amend this to read: "The results of the experiment are shown in Figures 9 to 12."

4. The definition of the key technology "Discriminative Deep Belief Networks" in the paper is inconsistently stated in the title and the main text. Please correct it according to the actual situation.

It is better to polish the written English by a native speaker.

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 5 Report

The paper presented the unmanned robot indoor positioning and navigation method with pseudolite in non-line-of-sight environments, and researches a Pseudolite fingerprint positioning method with the Discriminative Depth Belief Networks (DDBN) for positioning indoors. The test results of the proposed DDBN positioning method are better than that with the PSO-ANN, M-WKNN, and traditional WKNN algorithms.

Some comments are as follow:

1.     The Discriminative Depth Belief Networks (DDBN) is novel.

2.     The hardware and experiment performance are interesting to the readers.

Minor revisions:

3.     The last line of Page 2，Deep Belief Neural Networks (DDBN) should be “ DBNN”? Although your paper title is “…… Based on Discriminative Depth Belief Networks“. Why do you use the different acronym? Please check the whole paper.

4.     In Keywords，” carrier noise density” should be “Carrier noise density”.

Author Response

Please see the attachment.

Author Response File: Author Response.docx