An Optical Camera Communication Using Novel Hybrid Frequency Shift and Pulse Width Modulation Technique for Li-Fi

Round 1

Reviewer 1 Report

The manuscript approaches the issues associated to the Optical Camera Communications and claim to provide a novel hybrid frequency shift and pulse width modulation technique for LiFi. The positive aspects related to this work are based on the fact that the authors implement an OCC system and use it to evaluate the proposed modulation technique.

On the downside, the manuscript does not have a form that recommends its publishing. In many cases, the manuscript in written in a negligent manner. It seems that the authors skipped some of the proofreading steps, and they attributed these tasks on to the evaluators. Figures are introduced in a group and then, as the reader covers the manuscript, he has to get back to the Figures. In my opinion, the technical novelty provided by this work is limited, and thus, I do find the new modulation technique useful for OCC developers. I also found the article too long with respect to its content importance. In my opinion, this works brings limited new information to the OCC area, and in such cases, the 25 pages are too much. The results section is quite confusing. In theory, Bit Error Ratio is not expressed in percentage, but as a subunit number, for example 2.3x10^-3. Also, BER/bps is not a common metric. I believe that the authors have made same confusions here.

Other observations:

• Line 2: electronic components; it becomes -> the “;” should not be there
• Line 10: “photodiodes are observed to be inexpensive and provide low complexity implementation but have very short operational distance and are more sensitive to noise.” This information is not correct/misleading. The use of PIN photodiode enables tens of meters’ distances, whereas there are solutions to improve noise resilience.
• Lines 22-35: This section provide the same info as the abstract. Consequently, the abstract should be reorganized in order to be more attractive, and to provide new and more important elements.
• Lines 30-32: “Photodiodes are inexpensive and provide low complexity implementation for VLC as the modulation techniques are based on light intensity. In this case, achieving higher data rates is constrained to the response time of the photodiode [ 1 ][2 ][3 ].” The info is correct, but again misleading: data rate is determined by the PD response time, but PIN PD can provide response times as low as few ns. In the section above, it seems that the PD response time limits data rate performances. This could be true, but only when data rates higher than few Gb/s are envisioned.
• Page 2: Fig. 1 has a poor quality.
• Page 2: “Optical Camera Communication(OCC) is considered one possible solution towards achieving a higher, more stable data rate in VLC by utilizing the camera’s properties, computational capacity of mobile phones, and chromaticity of the light[6].” This info is false/ misleading. There are numerous VLC systems based on PIN PD that provide Gb/s data rates. OCC camera cannot provide similar data rates. Nevertheless, it is true that in certain applications (i.e. vehicular applications), OCC systems developed based on fast camera receivers (with 1000 fps for instance) can have better performances as they enable higher SNR.
• Lines 75-76: “Methods such as Background Subtraction, Polynomial regression, and Moving average.” This sentence is incomplete.
• Line 85: a comma is missing after “Thus”
• Line 106: when providing BER and distance, it would be useful to also provide data rate.
• Overview of Section 2: In section 1, the authors mention several times that PD VLC systems are not adequate due to their relatively low data rate. Nevertheless, in Section 2, the authors do not provide any OCC data rate example to demonstrate that OCC is superior in data rate. Off course, this cannot be demonstrated because it is not true. The authors should rethink the OCC motivation, focusing on the wide utilization of mobile phones that could enable OCC. Also, they should provide the data rates achieved by some of the current OCC systems.
• Line 192-193: “Since all the demodulation techniques will first have to perform some initial processing, a standard preprocessing function is implemented as shown in 1”. It is not clear what is 1. A figure, a table? Most probably the authors refer to Algorithm 1, but this should be clearly stated. The same for most of the figures, and algorithms.
• There are quite a lot misspelling errors, missing words, sentences that need to be rephrased. When few such errors are found, the reviewer can point them to the authors. However, the reviewer’s role is not to perform a manuscript spell check. This should have been done by the authors, prior to sending the manuscript for evaluation.
• The visibility of Figure 10, 16, 17 should be improved.

Author Response

Thank you for the suggestions and review comments. 

Summary of Major  Changes:

Figures 4,7,8 and 10 are removed from the original manuscript. Figures 1,2,5 and 12 are modified and quality has been improved in the new manuscript. Figure 8 is added in the new manuscript to show the experimental setup. Description of thresholding techniques under each modulation technique is reduced and merged under the decoding section of the corresponding modulation technique. Repeated function calls from algorithms are removed and explanations are added. Few sections are rewritten to reduce the total number of pages from 25 to 20. Grammatical mistakes and typos are corrected to the best of our knowledge.

Response to Reviewer 1 Comments:

Reviewer Comments: The manuscript approaches the issues associated with the Optical Camera Communications and claims to provide a novel hybrid frequency shift and pulse width modulation technique for LiFi. The positive aspects related to this work are based on the fact that the authors implement an OCC system and use it to evaluate the proposed modulation technique.

Point 1: On the downside, in many cases, the manuscript is written in a negligent manner. It seems that the authors skipped some of the proofreading steps, and they attributed these tasks to the evaluators. 

Response 1: Once again the proofreading is done and track changes are used to keep track of the changes made. 

Point 2: Figures are introduced in a group and then, as the reader covers the manuscript, he has to get back to the Figures. In my opinion, the technical novelty provided by this work is limited, and thus, I do find the new modulation technique useful for OCC developers. 

Response 2: Figures and descriptions are aligned properly in the manuscript. [Fig. 2,3 - Page 4, Fig. 4,5 - Page 5, Alg.1-Page 6]

Point 3: I also found the article too long with respect to its content importance. In my opinion, this work brings limited new information to the OCC area, and in such cases, the 25 pages are too much. 

Response 3: The size of the paper has been reduced to 20 pages by optimizing the content by removing repetitive information and rewriting paragraphs. Few images have also been edited to provide precise information.  

Point 4: The results section is quite confusing. In theory, the Bit Error Ratio is not expressed in percentage, but as a subunit number, for example, 2.3x10-3. Also, BER/bps is not a common metric. I believe that the authors have made some confusion here.

Response 4: The ratio of number of error bits at receiver to total number of transmitted bits is used for evaluating OOK, BFSOOK and HFSPDM. Bit rate has to be kept constant for comparing BERs, however,  OOK can accommodate multiple bits in a single video frame and thus operates at a very high bit rate compared to later two. The minimum bit rate achievable by OOK without flickering is many times higher than maximum bit rate achievable by BFSOOK and HFSPM. Thus, in an attempt to have a common comparison metrics ratio of BER to Bit rate is considered. 

Point 5: Line 5. Electronic components; it becomes -> the “;” should not be there

Response 5: The sentence has been changed as per the suggestion.

 

Point 6: Line 10. “photodiodes are observed to be inexpensive and provide low complexity implementation but have very short operational distance and are more sensitive to noise.” This information is not correct/misleading. The use of PIN photodiode enables tens of meters’ distances, whereas there are solutions to improve noise resilience.

Response 6: The statement has been changed to align with the motivation of proposed work. The changed statement is “Photodiodes are inexpensive and provide low complexity implementation, but their adoption requires modifying existing devices to house dedicated sensors.”

Point 7: Lines 22-35: This section provides the same info as the abstract. Consequently, the abstract should be reorganized in order to be more attractive and to provide new and more important elements.

Response 7: Necessary changes are made to the abstract based on suggestions. 

Point 8: Lines 30-32: “Photodiodes are inexpensive and provide low complexity implementation for VLC as the modulation techniques are based on light intensity. In this case, achieving higher data rates is constrained to the response time of the photodiode [ 1 ][2 ][3 ].” The info is correct, but again misleading: data rate is determined by the PD response time, but PIN PD can provide response times as low as a few ns. In the section above, it seems that the PD response time limits data rate performances. This could be true, but only when data rates higher than few Gb/s are envisioned.

Response 8: The statement is not modified as it conveys the intended information as per the cited works.

Point 9: Page 2: Fig. 1 has a poor quality.

Response 9: The figure has been slightly modified and the quality of the figure is also enhanced.

Point 10: Page 2: “Optical Camera Communication(OCC) is considered one possible solution towards achieving a higher, more stable data rate in VLC by utilizing the camera’s properties, computational capacity of mobile phones, and chromaticity of the light[6].” This info is false/ misleading. There are numerous VLC systems based on PIN PD that provide Gb/s data rates. OCC cameras cannot provide similar data rates. Nevertheless, it is true that in certain applications (i.e. vehicular applications), OCC systems developed based on fast camera receivers (with 1000 fps for instance) can have better performances as they enable higher SNR.

Response 10: The motivation of the proposed work is modified to align with the ease of integration and requirement of dedicated hardware. Thus, as per the suggestion, mobile phones and any other existing camera based systems can be easily adapted to achieve OCC. [Line 35] 

Point 11: Lines 75-76: “Methods such as Background Subtraction, Polynomial regression, and Moving average.” This sentence is incomplete.

Response 11: The sentence has been corrected. The new sentence is  “...provided an overview and performance evaluation of three thresholding methods, namely as  Background Subtraction, Polynomial Regression, and Moving Average.”[Line 79-80]

Point 12: Line 85: a comma is missing after “Thus”

Response 12: Comma is added to the sentence.

Point 13: Line 106: when providing BER and distance, it would be useful to also provide data rate.

Response 13: Data rate is added. “...BER of 3.8x10^-3 for the distance 0f 0.8m and data rate of  10.74 kb/s”[line 109]

Point 14: Overview of Section 2: In section 1, the authors mention several times that PD VLC systems are not adequate due to their relatively low data rate. Nevertheless, in Section 2, the authors do not provide any OCC data rate example to demonstrate that OCC is superior in data rate. Of course, this cannot be demonstrated because it is not true. The authors should rethink the OCC motivation, focusing on the wide utilization of mobile phones that could enable OCC. Also, they should provide the data rates achieved by some of the current OCC systems.

Response 14: The section is modified as per the suggestion.

Point 15: Line 192-193: “Since all the demodulation techniques will first have to perform some initial processing, a standard preprocessing function is implemented as shown in 1”. It is not clear what is 1. A figure, a table? Most probably the authors refer to Algorithm 1, but this should be clearly stated. The same for most of the figures, and algorithms.

Response 15: All the references to the figures, tables, and algorithms are updated as per the suggestion.

Point 16: There are quite a lot of misspelling errors, missing words, and sentences that need to be rephrased. When few such errors are found, the reviewer can point them to the authors. However, the reviewer’s role is not to perform a manuscript spell check. This should have been done by the authors, prior to sending the manuscript for evaluation.

Response 16: The manuscript has been revised completely to address the shortcomings. 

Point 17: The visibility of Figures 10, 16, and 17 should be improved.

Response 17: Figure 10 is removed and replaced with its text information. The quality of Fig. 16 and FIg.17 is improved.

 

Reviewer 2 Report

Dear Authors,

 The content of your article fits perfectly with the scope of Computation journal. One reason is that a key research topic involves data communication issues between devices in the new field of visible light VLC. This is very important for solutions to a newer purpose for the existing components.

The authors proposed a novel hybrid HFS-PDM type modulation technique. It is relevant and interesting.

The above-mentioned goal was based on the 18 publications analysed in the initial sections of the article.

The results of experiments carried out with appropriate modulation techniques are presented and explained. The authors proved that the approach proposed by them is much more effective.

The paper contains some new data.

The paper is presented in logical way and overall written well.

The content of the Conclusion is consistent with the evidences and arguments presented and addresses the main question asked. There is a reference to limitations in the article.

 Comments and Suggestions for Authors

1.   It would be best to clearly identify which of the previous works by all authors constitute the foundation of the work presented in this article.

2.   The study considered comparing the performance of the proposed technique with On-Off Keying (OOK) and Binary Frequency Shift-OOK (BFSOOK). What other alternatives were there?

3.   To what extent are the obtained results scalable, i.e. transferable to the higher frequency spectrum?

4.   Typos. The following typos were noted in the article: e.g., please to improve the readability of the contents in Figures 4, 6 and 7. 

Author Response

Thank you for the suggestions and review comments. 

Summary of Major  Changes:

Figures 4,7,8 and 10 are removed from the original manuscript. Figures 1,2,5 and 12 are modified and quality has been improved in the new manuscript. Figure 8 is added in the new manuscript to show the experimental setup. Description of thresholding techniques under each modulation technique is reduced and merged under the decoding section of the corresponding modulation technique. Repeated function calls from algorithms are removed and explanations are added. Few sections are rewritten to reduce the total number of pages from 25 to 20. Grammatical mistakes and typos are corrected to the best of our knowledge.

Response to Reviewer 2 Comments:

Reviewer Comments: The content of the article fits perfectly with the scope of the Computation journal. One reason is that a key research topic involves data communication issues between devices in the new field of visible light VLC. This is very important for solutions to a newer purpose for the existing components. The authors proposed a novel hybrid HFS-PDM type modulation technique. It is relevant and interesting. The above-mentioned goal was based on the 18 publications analyzed in the initial sections of the article. The results of experiments carried out with appropriate modulation techniques are presented and explained. The authors proved that the approach proposed by them is much more effective. The paper contains some new data. The paper is presented in a logical way and overall written well. The content of the Conclusion is consistent with the evidence and arguments presented and addresses the main question asked. There is a reference to limitations in the article.

 

Point 1: It would be best to clearly identify which of the previous works by all authors constitute the foundation of the work presented in this article.

Response 1: The works mentioned under section 2 provided insights on basics of OCC, existing modulation techniques and relation between the camera parameters and corresponding noise. Each work presented has shaped our motivation and guided towards implementation and evaluation of the proposed technique. Especially work presented by Rajagopal, N. et.al.[14] on OOK and Van Thang, N. et.al.[16] on BFSOOK were used for conducting experiments. However, works presented in [12] and [13] helped in shaping the thresholding technique for our proposed modulation. Through these and other works mentioned in section 2, it was identified that there is a need for low cost, ready to use, low complexity and robust OCC modulation.

Point 2: The study considered comparing the performance of the proposed technique with On-Off Keying (OOK) and Binary Frequency Shift-OOK (BFSOOK). What other alternatives were there?

Response 2: There are other modulation techniques that use properties of polarized light, chromaticity of light, spatial arrangement of transmitter, and 2D LED matrix. However OOK and BFSOOK are basic modulation techniques that have similar working principle and implementation complexity. The proposed technique aligns more with these modulation techniques compared to others mentioned before in terms of working principle and complexity. 

Point 3: To what extent are the obtained results scalable, i.e. transferable to the higher frequency spectrum?

Response 3: The proposed system uses the duration of frequencies to modulate the signal. Thus, the technique can be extended to higher frequencies as well. However, it must be noted that the difference between the frequencies should be distinguishable enough such that it can be identified using a rolling shutter camera. Fig. 10 and Section. 4.3.3.,  provide details on the constraints of the proposed technique.

Point 4: Typos. The following typos were noted in the article: e.g., please to improve the readability of the contents in Figures 4, 6, and 7. 

Response 4: The suggested typos have been corrected and the quality of the figures is enhanced.

Reviewer 3 Report

The authors have investigated an optical camera communication technique to achieve high accuracy using a hybrid frequency shift and pulse width modulation. The performance of the optical camera communication technique is compared with the OOK and the binary frequency shift-OOK. The manuscript is interesting and useful for the design and application of optical wireless systems. The paper is acceptable to be published in Computation, provided the following issue can be addressed

 

1. Some abbreviations should be clarified when they appear for the first time.
2. Add details in the table to describe the physical parameters of the considered transmission setup
3. Improve the quality of Figure 7.
4. Add some discussions on the state of art on optical wireless communication techniques to improve the performance of the considered system

See e.g.

W Hu et al., High-dimensional feature based non-coherent detection for multi-intensity modulated ultraviolet communications, [IEEE] Journal of Lightwave Technology, 2022.

H Matsunaga et al., I/Q demodulator based optical camera communications, IEEE Photonics Journal, 2022.

Author Response

Thank you for the suggestions and review comments. 

Summary of Major  Changes:

Figures 4,7,8 and 10 are removed from the original manuscript. Figures 1,2,5 and 12 are modified and quality has been improved in the new manuscript. Figure 8 is added in the new manuscript to show the experimental setup. Description of thresholding techniques under each modulation technique is reduced and merged under the decoding section of the corresponding modulation technique. Repeated function calls from algorithms are removed and explanations are added. Few sections are rewritten to reduce the total number of pages from 25 to 20. Grammatical mistakes and typos are corrected to the best of our knowledge.

Response to Reviewer 3 Comments:

Reviewer Comments: The authors have investigated an optical camera communication technique to achieve high accuracy using a hybrid frequency shift and pulse width modulation. The performance of the optical camera communication technique is compared with the OOK and the binary frequency shift-OOK. The manuscript is interesting and useful for the design and application of optical wireless systems. The paper is acceptable to be published in Computation, provided the following issue can be addressed

Point 1: Some abbreviations should be clarified when they appear for the first time.

Response 1: All the abbreviations are clarified as per the suggestion.

Point 2: Add details in the table to describe the physical parameters of the considered transmission setup

Response 2: The table mentioning experimental setup details has been added as per the suggestion.

Point 3: Improve the quality of Figure 7.

Response 3: Figure 7 is removed and more explanation is written instead.

Point 4: Add some discussions on the state of art in optical wireless communication techniques to improve the performance of the considered system

Response 4: The main motivation for the proposed work was to keep the implementation complexity minimal and to build a ready-to-use system without any special requirement on the receiver side. Although there exist various state-of-the-art systems that focus on specific aspects such as BER, Distance, and Complexity, it was observed that most of these systems use specialized hardware, thus incomparable to the proposed system. Out of the remaining works that align with our motivation, most of them have used transmitters as a beacon device for positioning. All the selected papers for comparison are discussed under Section 2 Related Work.

Reviewer 4 Report

This paper proposes to mix frequency shift and pulse width modulation is proposed for Li-fi. The idea is simple and some improvements compared to previous schemes have been shown. I have some comments for the authors in order to improve the manuscript more.

 

1. The length of the manuscript is too long. Some parts are redundance and should be shortened to concise the text and made the manuscript more readable. For example, the parts mentioned in OOK and BFSOOK schemes should be significantly reduced.

2. Instead of introducing OOK and BFSOOK, the FS and PDM parts should be introduced.

3. English expression needs to be improved, some typos and errors should be corrected. For example, there is an error in lines 131, 349.

4. The TestBed Setup section should be described in a more detailed table with all parameters such as transmission distance, noise power, etc. A full picture of the transmission scenario including both transmitter and receiver should be shown.

5. Authors should highlight the contributions of this paper in more detail. The objectives in 45-47 are too short and general.

6. In my opinion, the terms OOC, VLC, and Li-fi are similar in this context. So authors should consider using only one or two of those terms throughout the manuscript to avoid confusion for readers.

7. Figure 23 shows that BFSOOK is better than the proposed scheme in most cases under the same bit rate. So what is the advantage of the proposed scheme at the same bit rate? If the computation is an advantage, the complexity cost should be shown and compared in detail.

Author Response

Thank you for the suggestions and review comments. 

Summary of Major  Changes:

Figures 4,7,8 and 10 are removed from the original manuscript. Figures 1,2,5 and 12 are modified and quality has been improved in the new manuscript. Figure 8 is added in the new manuscript to show the experimental setup. Description of thresholding techniques under each modulation technique is reduced and merged under the decoding section of the corresponding modulation technique. Repeated function calls from algorithms are removed and explanations are added. Few sections are rewritten to reduce the total number of pages from 25 to 20. Grammatical mistakes and typos are corrected to the best of our knowledge.

Response to Reviewer 4 Comments

Point 1: The length of the manuscript is too long. Some parts are redundant and should be shortened to concise the text and make the manuscript more readable. For example, the parts mentioned in OOK and BFSOOK schemes should be significantly reduced.

Response 1: The size of the paper has been reduced by removing repetitive information and rewriting paragraphs. Few images have also been edited to provide precise information. 

 

Point 2: Instead of introducing OOK and BFSOOK, the FS and PDM parts should be introduced.

Response 2: Precise explanation of Frequency Shift Keying and Pulse Duration Modulation is added[Line 152-155]. The sections containing OOK and BFSOOK-related information are reduced but not completely removed since we wanted to convey implementational similarities and highlight the simplicity of the design of each technique. 

 

Point 3: English expression needs to be improved, and some typos and errors should be corrected. For example, there is an error in lines 131 and 349.

Response 3:  The suggested errors are corrected.

 

Point 4: The TestBed Setup section should be described in a more detailed table with all parameters such as transmission distance, noise power, etc. A full picture of the transmission scenario including both transmitter and receiver should be shown.

Response 4: As per the suggestion, Fig.8 is added to provide details of the experimental setup and transmission distance. Eq.15 and Eq.16 are added to provide details on how SNR was estimated. The values for Camera Brightness, Contrast, Exposure, and Ambient Light Intensity are all added under the experiment setup subsection.

 

Point 5: Authors should highlight the contributions of this paper in more detail. The objectives in 45-47 are too short and general.

Response 5:  The reframed objectives are more detailed as per the suggestion: 

Obj. 1: To study and experimentally evaluate thresholding techniques for On-Off Keying(OOK) and Frequency Shift Keying(FSK) modulation schemes under OCC.

Obj. 2: To propose and implement Hybrid Frequency Shift Pulse Duration Modulation scheme for OCC.

Obj. 3: To evaluate the performance of the proposed technique based on implementational complexity, Bit Error Ratio(%), distance, and frames per bit.

 

Point 6: In my opinion, the terms OOC, VLC, and Li-fi are similar in this context. So authors should consider using only one or two of those terms throughout the manuscript to avoid confusion for readers.

Response 6: As per the suggestion OCC based Li-Fi is the term used throughout the manuscript. Li-Fi is the domain while OCC is a technique used at the receiver side to enable communication under Li-Fi.

 

Point 7: Figure 23 shows that BFSOOK is better than the proposed scheme in most cases under the same bit rate. So what is the advantage of the proposed scheme at the same bit rate? If the computation is an advantage, the complexity cost should be shown and compared in detail.

Response 7:  The proposed scheme HFSPDM requires a lesser number of frames as compared to BFSOOK. It uses 5 frames to decode a pair of 2 bits, however, BFSOOK requires 3 frames for 1 bit as highlighted in Table.3. Figure 19, conveys that the proposed technique performs better than OOK for higher distances. Thus, concluding that the proposed technique is suitable for applications that require low computation and high BER.

Round 2

Reviewer 1 Report

The authors have addressed all my comments. The paper is now significantly improved.

Reviewer 4 Report

Thanks for addressing all comments. I have no further comments.

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

Please see below comments for Editor.

Reviewer 2 Report

The standard measures to describe the performance of communications systems is done by the SNR per bit (Eb/N0) at the receiver. In chapter 3.5 it is stated that for high symbol rates the width of the stripes is decreased and making the decoding difficult. How is that effect related to the SNR (Eb/N0) at the receiver (optical sensor)?

Chapter 4.1: Due to long consecutive zeros or ones in the data stream of OOK modulation, synchronization in the receiver is very difficult or even impossible. Why do the authors claim that OOK modulation is one of the most popular transmission techniques? References?

Fig. 7: Why are the amplitudes of the detected signals at the edges of the image sensor (pixels <150, pixels > 330) smaller than in the center of the image sensor? Does the received energy / "1" bit change so quickly? Please explain

 

For OOK modulation and high SNR, the optimal decision threshold is half the amplitude of the "1" bit. All algorithms shown in Fig. 7 to define a decision threshold, presents a large deviation from the optimal half bit amplitude.  Why, please explain

The SNR at the receiver and the decision threshold are the main factors influencing the expected bit error rate. How were the algorithms presented for calculating the decision threshold checked? What is the influence on the expected bit error rate?

The paper describes a Novel Hybrid Frequency Shift and Pulse Width Modulation Technique for LiFi but the performance of data transmission was not checked by the achievable bit error rate (BER) in dependency of the presented algorithms.

 

The effects of the presented algorithms on the bit error rate of data transmission were not investigated. Therefore, the performance of the presented modulation method was not checked to a significant extent.