ICACT20210200 Question.1
Questioner: kawano.masahiro959@mail.kyutech.jp 2021-02-09 오후 12:56:58 |
ICACT20210200 Answer.1
Answer by Auhor kenta-suzuki@atr.jp 2021-02-09 오후 12:56:58
Chrome  Click!! |
| Thank you for your very interesting presentation. My name is Kawano, and I am a student at Kyushu Institute of Technology. I have two questions on page 8 of the slide.
Q1. In the two figures, when the scan rate is high, it looks like a clean straight line almost like a linear approximation. Is it possible to obtain good characteristics that are close to linear approximation even when the scan rate is low?
Q2. If the scan rate is high, the transmission interval of terminals is short, or the number of terminals is large, the delay of scan response will increase significantly. Does this mean it is having a big impact on traffic?
Thank you.
|
Thank you for listening and questions, Mr.Kawano.
Answer for Q1:
The scan response delay and its variance are almost proportional. So, these regression functions monotonically increasing and linear characteristics can be obtained regardless of the scan rate.
However, if the scan rate is low, the difference in characteristics is not large, and accurate estimation is difficult.(Appendix slide 18 shows that the network state estimator is not perfect.)
If the scan rate is low, the impact on the network to be scanned may not be large, but evaluation with various network conditions is necessary.
Answer for Q2:
In each case of large amount of background traffic, collision between scan traffic and background traffic will occur frequently because of large transmission opportunity of the background traffic. So, if scanning when the delay is large, the impact on the background traffic will be large. |
ICACT20210200 Question.2
Questioner: yung.chien.shih@gmail.com 2021-02-09 오후 9:34:02 |
ICACT20210200 Answer.2
Answer by Auhor kenta-suzuki@atr.jp 2021-02-09 오후 9:34:02
Chrome Click!! |
| Thank you for this interesting research. I have a question about network situation estimator on pages 5 to 7. Assume that a case has different number of terminals which with different transmission intervals of the background traffic. How about estimation result and error rate? |
Thank you for listening and questions.
In a similar situation, when estimating with a estimator trained all training samples of cases 1 and 2, the estimation accuracy may not be good.
This is because the scan response delay has a wide range and the mean value is not a typical value in the target network.
In order to deal with various network situation, we consider that these situations need to classify in detail in order to select properly by Network situation estimator.
Therefore, performance evaluations for with various network situation is necessary. |
ICACT20210200 Question.3
Questioner: yung.chien.shih@gmail.com 2021-02-09 오후 9:16:18 |
ICACT20210200 Answer.3
Answer by Auhor kenta-suzuki@atr.jp 2021-02-09 오후 9:16:18
Chrome Click!! |
| What is application scenario or key value of proposed method on real network environment? |
For a IoT service using a large number of wireless devices such as a environment sensing,
a maintenance such as monitoring security level using port scan information or confirming
the existence need to conduct remotely and automatically.
In this situation, when conducting the maintenance with the proposed method,
the remote traffic do not disturb the existing service as much as possible. |
ICACT20210200 Question.4
Questioner: mistake 2021-02-09 오전 10:18:59 |
ICACT20210200 Answer.4
Answer by Auhor kenta-suzuki@atr.jp 2021-02-09 오전 10:18:59
Chrome Click!! |
| |
|
ICACT20210200 Question.5
Questioner: kiwi7580@gmail.com 2021-02-10 오전 10:27:16 |
ICACT20210200 Answer.5
Answer by Auhor kenta-suzuki@atr.jp 2021-02-10 오전 10:27:16
Chrome Click!! |
| What is the specification of your system for implementating the proposed method? |
Thank you for listening and question. We assume the proposed method implement to existing network scanner such as masscan which have a function to set the fixed transmission rate(*1).
(*1) https://github.com/robertdavidgraham/masscan |
ICACT20210200 Question.6
Questioner: Yiile@outlook.com 2021-02-10 오전 10:33:14 |
ICACT20210200 Answer.6
Answer by Auhor kenta-suzuki@atr.jp 2021-02-10 오전 10:33:14
Chrome Click!! |
| Thank you for your presentation. Have the research and experiments mentioned in the article been tested in the actual environment? |
Thank you for listening and question. The research and experiments have been only tested by network simulator.
The parameters of this experiments are only part of the Wi-SUN standard. Testing the proposed method in the actual environment is necessary. |
