A Fusion Strategy for Vehicle Positioning at Intersections Utilizing UWB and Onboard Sensors

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Dear Authors,

Your manuscript is welll written and results are well supported by data and conclusion. some further reference would need to be included in the introduction.

Comments on the Quality of English Language

the authors should revise the article and correct some expressions

Author Response

Thank you for your affirmation of our paper. To follow your suggestion, we have added more references in the introduction section.

Reviewer 2 Report

Comments and Suggestions for Authors

Dear Authors,

Appreciate your efforts in introducing approach related to vehicle localization using UWB, which is one of the interesting and demanding area for the autonomous driving. Manuscript written well and there are very few concerns, addressing those would greatly benefit the manuscript in improving clarity. 

Comments for author File: Comments.pdf

Comments on the Quality of English Language

Minor spell check and corrections required. 

Author Response

Thanks very much for your review of our paper. Your opinions are a great encouragement to us. In this paper, we focus on NLOS mitigation of UWB through ARIMA-GARCH model. It’s more effective than the previous method, which had already published in Sensors. In the experimental section, the reference data are capture by a strategic-grade GNSS/INS integrated navigation system named NovAtel SPAN-CPT. It can provide centimeter-level positioning date for the comparisons.

Reviewer 3 Report

Comments and Suggestions for Authors

Only 35% of the references are from the past 5 years, authors are required to update the reference.

Keywords have to be sorted alphabetically.

Define all acronyms when using for the first time, for e.g. what is ITS in the first line of the abstract.

Please use a formal writing, for example, don’t use “techs” instead of techniques line 11.

How can this algorithm be applicable on worldwide level, the complexity of the proposed scenario does not represent the real complex scenario of cities, where the effect of NLOS is dominant.

Why you didn’t include the localization using GPS in your comparison since you propose something that is supposed to be better than the GPS.

How is the performance of the proposed method when the UWB signal is very weak?

This topic is now widely adopted by integrating the AI technology, I have not seen any work from you or any mention of this in the reference list.

The paper is somehow similar to the published paper “An Integrated IMU and UWB Sensor Based Indoor Positioning System” please make it clearer why this paper is different from other papers.

Can you please tell what the velocity of the vehicle was? And what is the impact of velocity on the accuracy of estimation? 

Author Response

1.Thank you for your constructive suggestions. We had added more references in the introduction section.

2.We followed your advice, now the keyworks are sorted alphabetically.

3.We check the paper carefully. Now, all acronyms have defined when using for the first time.

4.Thank you for your careful reminder. We have changed the informal expression in our paper.

5.We tried our best to find a scenario to represent the real complex scenario of cities. However, the site for the ground test is limited. In this paper, we test the proposed algorithm at a typical urban intersection. Your opinion is very important, we will test the algorithm in more complex scenario in the future.

 6.In the ground test, we focused on verifying the effectiveness of the AIMM algorithm against the IMM method. Either the IMM or the AIMM method can achieve better performance than the GPS, in view of this, the comparison of GPS is omitted.

7.The max working distance of UWB module in this paper is about 200m. However, the range accuracy deteriorates when the distance between two UWB nodes is over 100m. The algorithm performs unsatisfactorily when the UWB signal is very weak.

8.Your advice is extremely valuable. As a matter of fact, we also start the research integrated with AI technology. We may exhibit our progress in the following papers.

9.We had revised our paper to make the novelty part of our work clearer. we believe the most novelty part is the fusion strategy: an adaptive IMM algorithm is developed to realize global fusion, in which the low-cost onboard sensors including electronic compass, wheel speed sensors and MEMS-INS are utilized to be fused with UWB. Compared with previous papers, the proposed AIMM model can adjust the model probabilities to make them more suitable for current driving conditions and thus leads to a better positioning performance.

10.In our field test, the velocity of the vehicle is about 0 to 60km/h. It’s an appropriate speed range for a vehicle at urban intersections. According to the experimental result, the velocity may not impact the accuracy of estimation.

We deeply appreciate your consideration of our manuscript, Thank you and best regards.