ICACT20230246 Question.1
Questioner: myy98@mail.sdu.edu.cn 2023-02-20 오후 7:08:26 |
ICACT20230246 Answer.1
Answer by Auhor hbw017@mail.sdu.edu.cn 2023-02-20 오후 7:08:26
Chrome  Click!! |
| Why is channel estimation in the DD domain rather than in the time-frequency domain? |
Thank you for your question. Because in the scene of high-speed vehicle movement, Doppler frequency shift and time delay will occur. DD domain can effectively reduce the impact of both on signal transmission, while the signal is greatly disturbed in time-frequency domain. |
ICACT20230246 Question.2
Questioner: myy98@mail.sdu.edu.cn 2023-02-20 오후 7:08:35 |
ICACT20230246 Answer.2
Answer by Auhor hbw017@mail.sdu.edu.cn 2023-02-20 오후 7:08:35
Chrome Click!! |
| What are the advantages of DCNN compared with CNN? |
Thank you for your question. We introduce residual path and HDC on the basis of CNN, which can not only effectively avoid the problem of CNN information loss, but also eliminate the noise characteristics trained by the network, so as to significantly improve the channel estimation accuracy of OTFS system. |
ICACT20230246 Question.3
Questioner: yjk1425@163.com 2023-02-20 오후 7:08:43 |
ICACT20230246 Answer.3
Answer by Auhor hbw017@mail.sdu.edu.cn 2023-02-20 오후 7:08:43
Chrome Click!! |
| Are there any other traditional methods for channel estimation considered in this paper? Why is the OMP algorithm used instead of others? |
Thank you for your question. We consider two traditional channel estimation methods: pilot-based channel estimation and OMP channel estimation. Because OMP algorithm shows good performance in traditional channel estimation research, so we choose OMP algorithm as the benchmark, and input the pre-training results into the neural network. |
ICACT20230246 Question.4
Questioner: lyh12208@163.com 2023-02-20 오후 7:09:02 |
ICACT20230246 Answer.4
Answer by Auhor hbw017@mail.sdu.edu.cn 2023-02-20 오후 7:09:02
Chrome Click!! |
| What are the pilot symbols and data symbols in OMP and what are their differences? |
Hello, thank you for your question. In the OMP channel estimation of the OTFS system, you need to insert pilots into the input data x in the DD domain to obtain the pre-estimation matrix. The data symbol represents the signal in DD domain, and the pilot represents the signal used for channel estimation. |
ICACT20230246 Question.5
Questioner: mjs9908@g.skku.edu 2023-02-20 오후 7:09:19 |
ICACT20230246 Answer.5
Answer by Auhor hbw017@mail.sdu.edu.cn 2023-02-20 오후 7:09:19
Chrome Click!! |
| hi, what kind of dataset did you use? |
Hello, thank you for your question. I'm using my own data set. According to the OTFS system I built, the real channel matrix and OMP algorithm pre-estimation matrix are generated, and these two are used as labels and inputs respectively. |
ICACT20230246 Question.6
Questioner: 21m107fl@stu.teikyo-u.ac.jp 2023-02-22 오후 8:34:07 |
ICACT20230246 Answer.6
Answer by Auhor hbw017@mail.sdu.edu.cn 2023-02-22 오후 8:34:07
Chrome Click!! |
| Thank you for your presentation! Could you briefly describe the future development of your research? |
Thank you for your question! In the future, we aim to study efficient and accurate channel estimation in high-speed vehicle scenarios. For the DCNN model proposed in the article, we are ready to change some model parameters and structure, and plan to build a network model with lower complexity. |
ICACT20230246 Question.7
Questioner: dingshengli@oru.edu 2023-02-22 오후 8:38:26 |
ICACT20230246 Answer.7
Answer by Auhor hbw017@mail.sdu.edu.cn 2023-02-22 오후 8:38:26
Chrome Click!! |
| Excellent work! What have you done to improve the accuracy of DCNN? |
Thank you for your question! We introduce Hybrid Dilated Convolution (HDC) in DCNN to expand the receptive field of convolution layers. At the same time, the residual path is introduced to eliminate the impact of noise in the data set during training, so as to improve the accuracy of channel estimation. |
ICACT20230246 Question.8
Questioner: office@icact.org 2023-02-28 오전 10:40:08 |
ICACT20230246 Answer.8
Answer by Auhor hbw017@mail.sdu.edu.cn 2023-02-28 오전 10:40:08
Chrome Click!! |
[Urgent]
Your final CRF full paper contains errors in the author's name, address, and e-mail address.
It must comply with the IEEE Template.
1. Only *, **, and *** must be used in the title segment.
2. Emails from all authors must be presented.
Currently, of the seven authors, only three are presented.
This is a serious disqualification for IEEE publication.
After quickly revising and uploading,
be sure to check the contents through the Author Webpage.
When this is done, please let us know immediately.
Albert, Publish Office of ICACT2023
|
|
