An Experimental Demonstration of MIMO C-OOK Scheme Based on Deep Learning for Optical Camera Communication System
Round 1
Reviewer 1 Report
The paper presents interesting experimental results related to implementation of deep learning of MIMO OCC. My comments:
1) Introduction part is not correct. References are not properly used and results mentioned are not from that papers. E.g. some examples
a) "LiFi technology can be deployed at a 10 m communication distance using a photodiode lens [19]."
however this paper does not include single mention of lens, nor 10m length span
b) "with 95% precision, are presented in [23]. However, multiple RGB cameras were applied in this study, and the cameras was parallel to one another."
However in this paper only single camera was used, and DGB LEDs based panel, not camera!! No 95% precision, but 99.4! And the same concept of MIMO which is biased by authors of this paper.
2) There are number of other papers on MIMO OCC from other teams, please improve it
3) Noise for OCC is not properly discussed. Please improve description how the pixel noise was considered in the measurement and how noise can change with different exposure
4) Improve conclusion by diving main results overview.
Author Response
Please check the attached file
Author Response File: 
Author Response.pdf
Reviewer 2 Report
Please find my comments attached as PDF file.
Comments for author File: Comments.pdf
Author Response
Please check the attached file
Author Response File: Author Response.pdf
Reviewer 3 Report
Section 1 Introduction
Rows 48,49: The claim “Visible Light Communication, Light Fidelity, and Optical Camera 48 Communication are three candidates for Optical Wireless Communication (OWC).” This is not exact, there are more than three candidates !
Shortcuts VLC, LiFi and OCC are not presented in a correct way !
Row 126: Why is used the YOLOv4 algorithm if the version 5 of this algorithm already exists ?
Section 3 System architecture
Figure 1: A lower part dedicated to the inputs contains only 1 (one) single camera. Is it correct if this figure presents multiple inputs ?
Row 219: There is a duplicate term “the Root mean square error (RMSE)”
Section 4 Implementation
Row 250: The variable a is not clearly defined. What exactly means “the mark and space amplitude” ? Is it a difference between amplitudes or is it valid for both amplitudes (mark, space) ?
Row 268: Bad number of Figure ! The devices used in this experiment are shown in Figure 2 5.
Rows 291,292: Variable incorporated into the formula (3) should be ITALIC !
Row 350,354: The value N must be specified more precisely. What exactly means N ? Is there any relationship between a result of the equation (7) and a number of repetition of DS ?
Author Response
Please check the attached file
Author Response File: Author Response.pdf
Round 2
Reviewer 2 Report
In case the paper is accepted, the authors should take into account the following comments:
L 53: delete the "E" at the end of the line
L 86/87: ... a photodiode-based ultra-high speed ...
L124/125: ... the mobility effect was not considered.
L 209/210: ... But the matched filter method did not perform well in a mobile environment, which ...
L 224: .. In this paper, a root mean square error (RMSE) approach was applied to ...
L 257/258: ... each symbol comprises one bit ...
L 292: ... environment, but it does not depend on ...
L 327: ... we can improve the SNR ...
References: Format and style of the items should satisfy the guidelines of the journal. It is suggested to take a published paper for comparison purposes.
Author Response
We already revised based on your comment.
please check the attached file
Author Response File: Author Response.pdf
