Round 1

Reviewer 1 Report

The authors investigated the hot topic of radar signal classification in the Radar Warning Receiver (RWR). According to the text of the article, it is possible to make some comments and suggestions:

1. The classification of radar signals is highly dependent on the shape of the pulse. Radars can use various pulse shapes: UWB, Gaussian pulse, modified Hermite polynomial shapes, raised cosine pulse shapes, etc. The classification of radar signals is highly dependent on the shape of the pulse. How will the shape of the radar pulses affect the classification of the radar signals? For which pulse shape did the authors conduct the Simulation experiment?
2. Line 288. There are 105 seconds and 140 seconds in the text. In Figure 5, the x-axis is time in microseconds. So which is correct: seconds or microseconds?
3. Do the authors have any explanation for the sharp increase (more than two times) in the duration of pulse processing with an increase in the number of processed pulses? What is the reason for the appearance of numerous narrow peaks in Figure 6?
4. The authors should be more precise about the terms used: accuracy, rate, accuracy rate. Lines 325-327. For radars 24 and 51 - recognition accuracy, and for radars 43, 41 and 44 - recognition rate? Accuracy and Rate are two different concepts placed in the same Table 5.
5. Line 281. Radar (No. 54) is indicated, and in Table 3 radar 51 is indicated.
6. In figure 5, radar 44 is indicated in the legend. Maybe it is radar 42, as in table 3?

Author Response

This document is divided into two parts. In part 1, all actions that we have made to our previous manuscript are listed. Part 2 is the response to peer reviewers’ comments.

1. List of actions

LOA1: Standardize the terms "real-time", "accuracy", "rate", and "accuracy rate" in the text.

LOA2: A detailed explanation of the relevant experimental conditions is supplemented.

LOA3: The parts of the text where the content of the expression is unclear have been revised.

LOA4: Some non-standard simulation diagrams were re-simulated and replaced.

LOA5: After checking all measurement formulas, make corrections to irregularities.

LOA6: Corrected typos and unsatisfactory formatting by reviewing the full text of the manuscript.

Everything we do with the manuscript is done through the revision function of the Microsoft Office Word software. In MOS word revision mode, all changes are highlighted with revision mode.

2. Response to reviewers’ comments

By summarizing the revising requirements which mentioned in the notification email from editorial office, there are six constructive comments have been made to our original manuscript. In this section, we will carefully reply to these comments one by one.

Comment 1: Influence of the shape of the pulse on radar recognition.

Response: Thanks for your good opinion. Following this suggestion, we rewrote lines 106-108. To facilitate evaluation of our revisions, we reproduce the revised abstract below.

The measured parameters form a list of characteristic parameter data in time series. Basic parameters include: Radio Frequency (RF), Time of Arrival (TOA), Pulse Width (PW), Pulse Amplitude (PA), Angle of Arrinal (AOA), etc. The list of data is sent to the backend processor. The latter performs signal classification, identification and threat assessment based on a list of data.

In addition, the author would like to explain: This paper mainly analyzes the back-end processing unit of the alarm receiver. At this stage, the radar signal is processed by parameter measurement and digitization of the front-end receiver to form the characteristic parameters describing the radar signal. In the signal identification of the alarm device, the identification mainly relies on the characteristic parameters of the signal.

Comment 2: Regarding the unit issue of "us" and "s" on line 288 and the inconsistency of radar numbers.

Response: Your advice is greatly appreciated. As suggest, The author re-standardized the range unit and number, and revised the simulation diagram. To facilitate evaluation of our revisions, we reproduce the revised abstract below.

Radar working mode
Radar5	[3000,3600]	[500,700]	[10,54]
Radar 1	[5200,5500]	[300,350]	[5,10]
Radar 3	[9800,11000]	[25,40]	[1,7]
Radar 4	[9600,9900]	[400,1500]	[0.5,2]
Radar 2	[9500,9900]	[30,60]	[0.5,2]

This paper selects 977190 pulse data generated during 105 seconds to 140 seconds in the demonstration process. The data parameter distribution is shown in Figure 5.

Comment 3: What is the reason for the appearance of numerous narrow peaks in Figure 6.

Response: The authors rewrote the data processing flow and explained that Figure 6 shows the timing of the pulses passing through the algorithm sequentially. The authors add the reason for the spike. To facilitate evaluation of our revisions, we reproduce the revised abstract below.

The parameter list of 1000 pulse features sequentially counts the time consumed by each pulse through the IFS algorithm. Statistics are obtained using the Monte Carlo method. The running time obtained by the simulation is shown in Figure 6.. As shown in the figure, the time required to calculate the single pulse is approximately 30 us, which can achieve the rapid identification recognition and alarm for the single pulse. However, the calculation time of the characteristic parameters for individual pulse data is long, and peaks appear in the graph.

Comment 4: Unification of the three terms "accuracy, rate, accuracy rate"

Response: This is good advice. The author carefully examines the section on the three terms in the text and unifies the terms. The changes are as follows.

• 350 lines: To reflect the IFS recognition results for each radar model, this study calculated recognition statistics, as shown in Table 5.
• 353 lines：Table 5 Recognition accuracy rate for radar
• 355 lines：Radar 3 has an approximate recognition accuracy of 85%,
• 365 lines：To address the low recognition accuracy of Radars 2, 3, and 4, the data were transmitted to the post-processing module,
• 373 lines：As shown in the table, the improved tri-training algorithm recognizes objects with an recognition accuracy of 88.91%, an increase of 38% compared to the IFS decision recognition algorithm used in the pre-processing module
• 405 lines：In future research, the initial selection of improved subsets can be examined to improve the stability and accuracy of the tri-training recognition accuracyrecognition rate.

Author Response File: Author Response.docx

Reviewer 2 Report

The paper deals with fuzzy methods for radar system recognition. In my opinion, the paper is readable, bringing original ideas and well organized. I have only a few minor notes and questions:

1. Have you performed experiments also with two and more radar signal with different power at the same time? How are the processing methods sensitive for such a disturbance?
2. You mentioned "real time" a few times in your paper. What does the real time mean? You used Matlab and probably a PC with Windows for the simulation. In my opinion such a configuration cannot be taking as a real time system.
3. You have tested your system with a few radar types, marked by numbers. The numbers (and radars) mentioned in the paper seems to be chosen from a bigger variety. Can you give some explanation to this?
4. Please check the figures - the text size in them is much smaller than in the plain paper text. From the formal point of view and paper readability I recommend to improve them.

Author Response

This document is divided into two parts. In part 1, all actions that we have made to our previous manuscript are listed. Part 2 is the response to peer reviewers’ comments.

1. List of actions

LOA1: Standardize the terms "real-time", "accuracy", "rate", and "accuracy rate" in the text.

LOA2: A detailed explanation of the relevant experimental conditions is supplemented.

LOA3: The parts of the text where the content of the expression is unclear have been revised.

LOA4: Some non-standard simulation diagrams were re-simulated and replaced.

LOA5: After checking all measurement formulas, make corrections to irregularities.

LOA6: Corrected typos and unsatisfactory formatting by reviewing the full text of the manuscript.

Everything we do with the manuscript is done through the revision function of the Microsoft Office Word software. In MOS word revision mode, all changes are highlighted with revision mode.

2. Response to reviewers’ comments

By summarizing the revising requirements which mentioned in the notification email from editorial office, there are six constructive comments have been made to our original manuscript. In this section, we will carefully reply to these comments one by one.

Comment 1: Regarding the influence of two or more radar signals of different powers on the experimental results.

Response: Thanks for your good opinion. In the data taken in this paper, there are situations where different radar powers are different. Following this suggestion, we rewrote lines 294. To facilitate evaluation of our revisions, we reproduce the revised abstract below.

This paper selects 977190 pulse data generated during 105 seconds to 140 seconds in the demonstration process. The data parameter distribution is shown in Figure 5. Among them: radar 1 has a total of 106,827 pulse data; radar 2 has a total of 94,043 data; radar 3 has a total of 128,610 data; radar 4 has a total of 605,650 data; radar 5 has a total of 20,700 data..

In addition, the author would like to explain: Signal dilution and signal sorting are completed in the previous processing link of signal identification. Then the alarm device completes the signal recognition for the radar signal in sequence according to the time series.

Comment 2: On the meaning of "real-time" and the contradiction between "real-time" and simulation with Matlab

Response: Your advice is greatly appreciated. As suggest, The author double-checked "real-time" and used "rapid" instead for the simulation part. And the simulation will be explained.. To facilitate evaluation of our revisions, we reproduce the revised abstract below.

This paper selects 977190 pulse data generated during 105 seconds to 140 seconds in the demonstration process. The data parameter distribution is shown in Figure 5.

 

Comment 3: What is the reason for the appearance of numerous narrow peaks in Figure 6.

Response: The authors rewrote the data processing flow and explained that Figure 6 shows the timing of the pulses passing through the algorithm sequentially. The authors add the reason for the spike. To facilitate evaluation of our revisions, we reproduce the revised abstract below.

(1) 312 lines: Since this paper does not have the conditions for the actual application of the algorithm, the MATLAB simulation platform is used to verify the rapid computing capability of the algorithm. The simulation software environment used in this paper is listed in Table 4,. The parameter list of 1000 pulse features sequentially counts the time consumed by each pulse through the IFS algorithm. Statistics are obtained using the Monte Carlo method. The running time obtained by the simulation is shown in Figure 6.and the running time obtained by the simulation is depicted in Figure 6. As shown in the figure, the time required to calculate the single pulse is approximately 30 us, which can be improved to achieve the rapid identificationreal-time recognition and alarm for the single pulse.

(2) 8 lines: Radar Warning Receiver (RWR) is one of the basic pieces of combat equipment necessary for the electromagnetic situational awareness of aircraft in modern operations that requires good rapidreal-time performance and accuracy.

(3) 12 lines: In the front-level information processing module, multi-attribute decision-making under intui-tionistic fuzzy information (IFS) is used to process radar signals with certain prior knowledge to achieve rapidreal-time performance.

Comment 4: Regarding radar types and numbers.

Response: This is good advice. The author has multiple radar parameters in the simulation, and randomly selects 5 radars of various types. To prevent ambiguity, the authors renumber the radars.

Several radar parameters are set in the simulation system. According to different radar types, 5 radar signal data are randomly selected., namely the early warning radar (radar 5), the guidance radar (radar 1), and the airborne radars (radar 2, radar 4, and radar 3). The range of attributes associated with each radar is listed in Table 3.

Table 3 Radar working mode attribute range

Radar working mode
Radar5	[3000,3600]	[500,700]	[10,54]
Radar 1	[5200,5500]	[300,350]	[5,10]
Radar 3	[9800,11000]	[25,40]	[1,7]
Radar 4	[9600,9900]	[400,1500]	[0.5,2]
Radar 2	[9500,9900]	[30,60]	[0.5,2]

Author Response File: Author Response.docx