Research on ELoran Demodulation Algorithm Based on Multiclass Support Vector Machine

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

This paper proposes a new demodulation algorithm leveraging a multiclass support vector machine (MSVM) for pulse position modulation (PPM) of eLoran signals. Through comprehensive experiments, the algorithm parameters are optimized, and the parallel comparison of different demodulation methods is carried out in various complex environments. The test results show that the MSVM algorithm is superior to traditional methods and other kinds of machine learning algorithms in demodulation accuracy and stability, particularly in high noise and interference scenarios. The paper has a certain degree of innovation. The suggested changes are as follows.

1.      In the feature vector construction section, the paper mentions the use of 16-dimensional feature vectors but does not provide a clear explanation for selecting this specific dimension. The choice of dimensionality can significantly impact the algorithm's performance. It is recommended that the authors explain the reasoning behind selecting 16 dimensions, including any empirical evidence, theoretical support, or comparisons with other dimensionalities, to help readers understand the rationale and applicability of this choice.

2.      The machine learning method in this paper needs to be trained with 90000 signals and applied after the training. How can this method be used in a practical receiver? The data characteristics of the receiver in practical application are different from those of the receiver in training. How will this affect the decoding performance?

3、In the experimental design, the amplitude of Lauret signal is absent, and quantization bits of the signal are not accounted for. In the experimental results, graphs (figures 11 and 14) and tables (table1,2, and 3) lacked key information about CWI frequency and amplitude. These parameters are the key parameters of the algorithm simulation and need to be explained in relevant chapters.

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 2 Report

Comments and Suggestions for Authors

It is of great value to study ELoran Demodulation Algorithm. The author proposes an algorithm for leveraging a multiclass support vector machine (MSVM) for pulzse position modulation (PPM) of eLoran signals. The authors have made some contributions to this topic. However, there are still many inadequacies in this manuscript. I hope the following comments can help to improve the manuscript. In my opinion, minor revision should to be done before this manuscript could be accepted for publication.

 

1. Although the current state of relevant research by many scholars has been presented, the introduction does not provide an overview of previous work in signal pulse modulation using machine learning to compare with the machine learning algorithms used in this paper.

 

2. The author could provide a brief explanation of α and β in the Formula 2.

 

3. In line 432：Does the use of different GRI values in the input signal affect the experimental results?

 

4. The conclusion section presents the results of the comparative analysis of the different experimental protocols，and it is proposed to briefly explain the limitations of signal demodulation based on this algorithm.

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 3 Report

Comments and Suggestions for Authors

The paper describes a novel demodulation method for the eLORAN system, which is shown to be superior to the previous demodulation methods. The paper is a very valuable contribution to the development of resilience of the eLORAN system and a good basis for the next research. The paper has good introduction and impressive number of references.

 

A few minor comments/questions:

Figs. 4 and 7 should be introduced after they are mentioned in the text;

Lines 192+: what are the parameters of the CWI used in Fig. 4 in relation to the wanted LORAN-C signal (e.g. amplitude/power, bandwidth etc.)? Is it the same CWI as in lines 437-440?

Fig. 5 - the axes need descriptions (even if they are exemplary); also, in Fig. 5b, the value of b is presented as 0 (special case) - the hyperplane should be shown in a more general way.

Comments on the Quality of English Language

Lines 34, 97, 119 - minor language errors.

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

All my comments have been addressed in the revised version, and I have no more questions.