Anti-Jamming Method and Implementation for GNSS Receiver Based on Array Antenna Rotation
Round 1
Reviewer 1 Report (Previous Reviewer 3)
Please review attachment
Comments for author File: 
Comments.pdf
Author Response
We are grateful for the advices and comments of all the reviewers and editors. The manuscript is revised according to the comments.
Because almost the whole introduction and the paper structure have been modified, track change” function is not used for huge revision. Instead, annotations are added to prompt the modification.
We’re glad to make further revisions based on the helpful comments of reviewers and editors.
Reviewer 2 Report (New Reviewer)
This paper utilized the direction sensitivity of the anti-jamming performance and proposed an anti-jamming method for array antenna receivers based on antenna rotation. A variable-step iteration algorithm based on gradient-descent then is given to determine the optimal rotation angle. From my point of view, major revision is needed for this paper before publishing. My suggestions for adjustments are as follows,
1. The English proofing is needed to improve the fluency of the text and to make it more comprehensible.
2. The introduction part is too long with too much redundant information. It needs to get into the main theme more quickly and concisely. For now, it almost lost its readability as a scientific paper. At the same time, commonsense knowledge does not need to be introduced in detail.
3. The writing format of the article is not standard and needs to be polished in-depth, such as in lines 45 and 46. Too many repeated abbreviations are used.
4. The font sizes in certain equations and figures are too small compared to the font size of the text.
5. Add more contributions and list your own contributions in the paper.
6. As mentioned in the discussion, the models, methods, and experimental settings in this paper are too simple to be presented. The parameter considerations are oversimplified. And the applicability of the model is relatively limited. But the solution is advanced to a certain extent.
7. This paper is hard to follow. The equations are not well written or well explained. Some symbols are not clearly explained. Certain inference processes are introduced too much. There is a lack of a block diagram of your work or a flowchart to show the algorithm.
8. For the last comment, I would suggest that if the work is not presented and explained well, even the best technical results are hard to be approved worthy.
Author Response
Please see the attachment.
Reviewer 3 Report (New Reviewer)
The manuscript presents an algorithm for minimizing the jamming power entering a GNSS receiver, using an antenna array with physical rotation capabilities. The target use case is an airborne receiver, subject to jamming from spatially separated sources.
Although the core algorithm for the optimization of the suppression capabilities of the antenna array may have potentials, several flaws can be recognized in the adopted hypotheses, addressed context and development of the study.
The English needs to be carefully revised, as several typos or unexpected expressions are present, e.g., “it’s” instead of “its”, “interferences” instead of “interference” (singular in English), “meats” instead of “meets”, “disables the receiver”, “equality harmful”, “there are a few literatures”, etc.
Major points to be revised:
1) Figure 1 is just qualitative, but the concept is well-known. The authors should either present a quantitative example taken from a realistic signal simulation (realistic signal and interference levels, cross-correlation peak as the result of an acquisition process etc.), or remove the figure.
2) In Figure 2 z_a should be pointing towards the Earth surface.
3) Page 3, line 73-74. The number of antenna elements could be limited to 3-4 in pure civil aviation installations.
4) Table 1 should also list the notation used on equation 9.
5) In Figure 3, the font size is too tiny, please increase it for readability.
6) One of the major assumptions of the work is the presence of an “External assistance” to obtain interference DOA, power, channel noise and antenna attitude. However, these are typically the most critical information to estimate and provide. The assumption of an “external assistance” able to provide the information with negligible delay, negligible error, ideal synchronization, perfect geometrical alignment and no conversion errors between reference frames is decisively too simplistic.
7) Page 4, line 148: what does “Q” represents? What is the meaning of “quality factor of the interference”?
8) Page 5, eq. (7). R_xx should be defined close to eq. (4).
9) Page 5, eq. (8). The construction of R_xx depends on too many factors assumed here as perfectly known (see point 6 before), but this is unrealistic.
10) Page 5, line 175. It appears that the receiver is assumed as stationary. But this is in contrast with the airborne scenario mentioned above, though. Please clarify.
11) Pages 5-6. The antenna array rotation is described. Nonetheless, I wonder how realistic it is the assumption of a physically rotating element in an airborne context. Furthermore, this rotating antenna is used for “only optimizing” the array direction during the flight. The gain offered by continuously rotating the array in the optimal direction is not convincing in the discussion of the authors. Another option, likely less expensive and far easier to maintain, could be to increase the number of sensors in a stationary array. It is expected at this point of the work that the benefit of rotating the antenna in the optimal direction reduces as the number of interference sources increases. A comparative cost-benefit analysis in this sense should be definitely addressed by the authors, otherwise the solution they propose appears not solid enough.
12) Page 6: The meaning and usability of the reported example is not clear: the authors choose a number of interference equal to the number of antenna elements (although on page 4, line 125, they claimed “This paper concentrates on the condition that K >= N”), equally spaced with respect Delta-phi = 10 deg. It seems quate a simple situation to discussion. Which is the performance at Delta-phi=0? And for K >> N? On page 7, line 230, which is the meaning of the sentence “the residual interference power has a similar change trend that can be taken advantage of in anti-jamming”?
13) Page 8, line 249. How it is possible to claim that “the noise power is close to zero”? The noise power cannot be reduced! The presence of the noise power can be assumed to be negligible when the gradient is computed.
14) Page 8, Line 257. Sentence “Covariance matrix Rxx( ) is constructed from the information of interference intensity, incident direction and rotation angle.” As written before, this is a very critical point. The sensitivity of the algorithm to the errors in the construction of Rxx should be assessed, because the source of such an information cannot be realistically assumed error-free.
15) Page 9, Section 3.3. This section appears to discuss about a very special and optimistic situation, in which the DOAs of the interference sources are symmetrical and their number is K==N. A very low-probability situation. Furthermore, the symmetricity was already evident before. Where is the information here? How general is the discussion?
16) Page 11, lines 347-352. This conclusion relies on the hypotheses of knowing: the number, the power levels and the angles of arrival (with respect to the array!) of different interference sources. As said, this hypothesis is too critical to be simplistically by-passed: it must be validated carefully.
17) Page 12, line 385. I do not see why the typical value for DeltaPsi is lower than 90 degrees. Why do the authors assume that the jammers are concentrated within 1 quadrant with respect to the aircraft trajectory? For example, especially when the aircraft is landing, the altitude is reduced, the probability of jamming exposure is the highest (either intentional or non-intentional) and the relative position between the jamming sources and the aircraft seems not to have a preferred direction. Furthermore, K=N=4 is not generic enough for an algorithm which claims to be robust for K>>N.
18) Figure 8. The starting point of the algorithm is not indicated, and the font size is too tiny.
19) Page 13, Section 4.3:
a. Monte-Carlo simulations with 3000 runs are introduced. Which parameters are varied in each run? DOA? Noise? Number and power of the sources?
b. Why does the optimum rotation angle not change? Are the simulations representative enough?
c. From Figure 10 I understand that the acquisition threshold is reached only for K=4. So is the algorithm useless for K>4? The discussion and the figure are not clear (or the algorithm is useless)
Author Response
We are grateful for the advices and comments of all the reviewers and editors. The manuscript is revised according to the comments.
Because almost the whole introduction and the paper structure have been modified, track change” function is not used for huge revision. Instead, annotations are added to prompt the modification.
We’re glad to make further revisions based on the helpful comments of reviewers and editors.
Please see the attachment for responses.
Round 2
Reviewer 2 Report (New Reviewer)
My comments have been addressed. Thank you. But still I suggest a serious proofreading and enhance the quality of the Figures.
This manuscript is a resubmission of an earlier submission. The following is a list of the peer review reports and author responses from that submission.
Round 1
Reviewer 1 Report
Comments for author File: Comments.pdf
Reviewer 2 Report
Some general comments:
- Language must have to be improved to a satisfactory level.
- In page 13, JSR appears to be too high for any GNSS signal to survive. JSR in the range of 30 to 45 dB would be enough to block reception of GNSS signal. There was no motivation or analysis presented on the choice of JSR.
- What is the reason for choosing 1268.52 MHz as the carrier frequency, not 1176.45 MHz or 1575.42 MHz or something else.
- The authors did not present any realistic simulation model where a GNSS receiver is present in the scenario. The discussion altogether lacks methodological analysis.
- The context of GNSS signal acquisition and tracking is only mentioned in the manuscript without offering any insight on how that practically get affected in the presence of multiple jammers.
- Motivation for only using AWGN as interferer is not explained. Usually commercial cheap jammers have individual jamming signatures (for example, chirp-like) that may have greater impact than AWGN.
- A realistic scenario with live data will be needed to back up the findings.
Comments for author File: Comments.pdf
Reviewer 3 Report
· The paper introduces Global Navigation Satellite System (GNSS) as providing positioned timing services for applications in transportation, marine fisheries, geo-hazards, and emergency rescues and many other applications
· However, the unintended or intentional interruption can easily jeopardize the receivers under complex EM environment threaten the weakening of the signal acquisition and tracking process, terming distributed interference as the most extremely harmful in this regards, declaring the importance of anti-jamming methods of the GNSS receivers
· Anti-jamming methods can be time domain, frequency domain, space and space-time domain methods among which the spatial based on the array antenna to suppress broad-band interference most effectively.
· The authors claim that with the development of jamming equipment and technology, the cost for increasing number of interference sources is relatively low, and the limitations for increasing the number of elements are huge.
· Describing the common anti-jamming algorithms of Power Inversion (PI) and Minimum Variance Distortion less Response (MVDR), both of these methods have limitations in terms of the output power and the type of filter used.
· Thus this paper proposes anti-jamming method for GNSS receivers based on antenna array rotation together with an efficient way based on gradient-descent iteration to realize the method.
· The Influence of antenna array attitude on Anti-jamming is explained in Fig. 2 indicating the antenna rotation and its effect, while the flow of the overall algorithm is given in Fig. 1, and In Fig. 3 shows the incident elevations of interference as -15°, and the azimuth angles are 30°, 10°, -30°, and -50° respectively, while the JSRs are set to 60dB.
· The selection of initial iteration values and the selection of spatial filter weights of PI algorithm under the mentioned conditions are given in Fig. 4.
· The simulation and analysis are explained in Fig. 5, and the relevant results in terms of the effectiveness of rotating antenna array algorithm, and those of the anti-jamming ability of PI algorithm and the proposed MVDR algorithm are compared as shown in Fig. 6.
· The relationship between efficiency improvements of the proposed method and interference intensities are given in Fig. 9
· The paper is concluded in Discussion which is a kind of Conclusion properly quantified, and reflecting what theoretical and simulation analyses have been promised.
Is the subject matter presented in a comprehensive manner?
I am convinced that the authors have done justice to support title “Anti-jamming Method and Implementation of GNSS Receiver-Based on Array Antenna Rotation”. The theory has been explained well, and the parameters of the research are clearly identified and presented in results of Fig. 6, Fig. 8 and Fig. 9. There is considerable of element for the technical body of as the paper has been presented with a meaningful outcome, the conclusion has been covered in Discussion which shows what has been promised in the Abstract.
Are the references provided applicable and sufficient?
The authors take support from thirty three (33) mostly old and branded references with few new, including none (0) from MDPI. I suggest to kindly remove references older than 2000.
Further, the Abstract needs be divided into Background/Motivation, Method, Results/Application, and Utility/Impact. The Discussion be converted into Conclusion though it is drawn based on the Results, making sure that what has been promised in Abstract is delivered in Conclusion.” I suggest that the authors rename Discussion as Conclusion.
Comments for author File: Comments.pdf
