Next-Generation IoT: Harnessing AI for Enhanced Localization and Energy Harvesting in Backscatter Communications

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

On the whole, the research of this manuscript has certain significance, but there are many problems that need further revision.

1. The overall academic value of the manuscript is not high. The content of the manuscript is generally simple. This manuscript has the low reference and usage value. The content of the manuscript is generally empty. The new methods are not truly proposed. It is recommended to further increase the content of manuscript to improve its overall value.

2. The content in Figure 7 is not well explained. The meanings represented on the vertical and horizontal coordinates in Figure 7 are not given There is no detailed explanation given as to why the mathematical expression for the vertical axis is so.

3. It is recommended to avoid using 'we' in the manuscript. The expression in the manuscript should be from the perspective of the third person.

4. The main text of the manuscript should first provide "XXX is as shown in the Table/Figure X", then the authors should provide the corresponding Table or Figure. Otherwise, it is easy to mislead readers and does not meet the requirements of academic articles.

5. The grammar in the manuscript is generally good, but in academic journals, the passive voice is generally the main voice. It is recommended to revise the manuscript to address this issue.

Comments on the Quality of English Language

The grammar in the manuscript is generally good, but in academic journals, the passive voice is generally the main voice. It is recommended to revise the manuscript to address this issue.

Author Response

Reviewer-1

On the whole, the research of this manuscript has certain significance, but there are many problems that need further revision.

 

1. The overall academic value of the manuscript is not high. The content of the manuscript is generally simple. This manuscript has the low reference and usage value. The content of the manuscript is generally empty. The new methods are not truly proposed. It is recommended to further increase the content of manuscript to improve its overall value.

 

Response: Thank you for your insightful feedback on our manuscript. We appreciate the time you have taken to review our work and share your thoughts. Based on your comments, we have given the manuscript a thorough revision. We have added more depth to the content to make it more substantial and informative. We have also taken a closer look at our methods to highlight their uniqueness and relevance more clearly. To strengthen the manuscript further, we have provided a more detailed explanation of our proposed localisation mechanism and testbed. We have also increased the number of references to strengthen the manuscript's foundation and utility for future research. We hope these changes have lifted the manuscript to meet the high standards of quality publication. Your feedback has been incredibly helpful in guiding these improvements, and we're truly grateful for it.

 

2. The content in Figure 7 is not well explained. The meanings represented on the vertical and horizontal coordinates in Figure 7 are not given There is no detailed explanation given as to why the mathematical expression for the vertical axis is so.

 

Response: In response to the reviewer's comment regarding the lack of explanation for the coordinates and the mathematical expression in Figure 7, we have revised the description to provide a clearer understanding. The horizontal axis (X-axis) in Figure 7 represents the target variable, which is the variable we are trying to predict or explain. The vertical axis (Y-axis) represents the output variable, the outcome of our model's predictions. The mathematical expression on the Y-axis, "Output = 0.987*Target + 0.08", is the linear regression equation derived from our model training. This equation is fundamental to understanding the relationship between the output and target variables. The slope (0.987) of this equation indicates how much the output variable is expected to increase when the target variable increases by one unit. The intercept (0.08) provides the expected value of the output variable when the target variable is zero. This equation was derived from the statistical analysis of our dataset and is crucial for interpreting the regression plots in Figure 7.

The newly added text is updated (in boldface) on pages 10 and 11 of the revised manuscript.

 

3. It is recommended to avoid using 'we' in the manuscript. The expression in the manuscript should be from the perspective of the third person.

 

Response: Thank you for your recommendation regarding using perspective in our manuscript. We appreciate your guidance on maintaining academic writing standards. In response to your suggestion, we have revised the manuscript to adopt a third-person perspective throughout the text. This change aligns the manuscript with the conventional academic writing style and enhances its formal tone. We are grateful for your insightful feedback, which has helped us improve the clarity and professionalism of our manuscript.

 

4. The main text of the manuscript should first provide "XXX is as shown in the Table/Figure X", then the authors should provide the corresponding Table or Figure. Otherwise, it is easy to mislead readers and does not meet the requirements of academic articles.

 

Response: We are thankful for your constructive feedback on the structure of our manuscript, particularly regarding the placement of references to tables and figures. Following your advice, we have revised the manuscript to ensure that the corresponding visual aid immediately follows each mention of a table or figure. This adjustment will undoubtedly make the manuscript more reader-friendly and align it with the standards of academic writing. Your input has been invaluable in enhancing the coherence and clarity of our work.

 

5. The grammar in the manuscript is generally good, but in academic journals, the passive voice is generally the main voice. It is recommended to revise the manuscript to address this issue.

 

Response: Thank you for your positive remarks on the grammar of our manuscript and your suggestion regarding the use of voice. We understand the importance of adhering to the conventions of academic writing, and in line with your recommendation, we have revised the manuscript to incorporate more passive voice constructions. This revision aims to align the manuscript with the formal style typically expected in academic journals. We are grateful for your guidance, which has been instrumental in refining the manuscript's linguistic style.

 

Reviewer 2 Report

Comments and Suggestions for Authors

The paper has presented a Potential for Future AI-enabled Localisation with Energy Harvesting in Backscatter Communications for IoT. It is meaningful in some extent. Here also has some issues should be addressed below.

1. It would be better present the paper with the third person，not the first person such as we.

2. The exactly objection is not clear in abstract, the highlights should be improved further.

3. What is the difference that applying backscatter communication in this paper with already existed research before, what is the innovation in this paper. It should be presented in further in introduction.

4. How did this paper realize the energy harvesting, the experimental should present with more clear.

5. The results and discussion is not clear, and in section 5 why methods were presented in this section, this section is the results according to the paper, so please revise them, the method and experimental scheme should clear presented.

6. The conclusion should include the extend of the future application or other.

7. The title why used potential? It may be improved with a accurately title.

Author Response

Reviewer-2

The paper has presented a Potential for Future AI-enabled Localisation with Energy Harvesting in Backscatter Communications for IoT. It is meaningful in some extent. Here also has some issues should be addressed below.

1. It would be better present the paper with the third person，not the first person such as we.

Response: Thank you for your valuable feedback regarding the writing style of our paper. We appreciate your suggestion to present the paper in the third person rather than the first person. This change enhances the formal tone and objectivity of academic writing. We will revise the manuscript accordingly to ensure it adheres to this more formal and widely accepted style in scientific literature..

2. The exactly objection is not clear in abstract, the highlights should be improved further.

Response: Thank you for your insightful feedback regarding the clarity of our abstract. We have revised the abstract to outline the objectives and highlights of our research more explicitly. The revised abstract now clearly states the challenges in complex RF environments for backscatter communications and localisation, particularly in IoT networks. It outlines our approach to address these challenges using a machine learning framework with K-Nearest Neighbors and Random Forest classifiers. The abstract also emphasises our key findings, including the high precision accuracy of our localisation method and the successful energy harvesting capabilities, which are crucial for IoT applications. This revised abstract provides a more precise and more comprehensive overview of our study, effectively setting the stage for the detailed exploration that follows in the paper.

The abstract has now been updated in the revised manuscript.

3. What is the difference that applying backscatter communication in this paper with already existed research before, what is the innovation in this paper. It should be presented in further in introduction.

Response: Thak you for your valuable comment. Our paper introduces several novel elements in response to your query regarding the distinction and innovation of our backscatter communication approach compared to existing research. Firstly, we incorporate advanced Machine Learning (ML) techniques for localisation within IoT networks, a methodology that has yet to be extensively explored in prior studies. This approach allows for more precise and dynamic localisation in variable environments, a key challenge in previous research. Secondly, integrating energy harvesting techniques within the backscatter communication framework is innovative. This dual focus on energy efficiency and effective data transmission in IoT devices presents a unique contribution to the field. In the introduction, we will elaborate on these innovations to more clearly delineate our work from existing literature.

4. How did this paper realize the energy harvesting, the experimental should present with more clear.

Response: Thank you for your feedback regarding the clarity of our experimental setup for energy harvesting. In our paper, energy harvesting is realised by deploying battery-less RF backscatter devices. These devices are strategically selected and positioned to optimise energy capture from available RF sources. In the revised manuscript, we will include a more detailed description of this setup, along with diagrams and specific data on energy capture efficiency, to provide a clearer and more comprehensive understanding of our experimental approach. This enhancement will ensure a better comprehension of the innovative aspects of our energy harvesting methodology.

 

 

 

 

 

 

5. The results and discussion is not clear, and in section 5 why methods were presented in this section, this section is the results according to the paper, so please revise them, the method and experimental scheme should clear presented.

Response: We appreciate your feedback on the clarity of the Results and Discussion section. In response to your concerns, we have thoroughly revised Section 5 to enhance clarity and coherence. The methods and experimental schemes have been distinctly outlined and separated from the results. We have also elaborated on the results, particularly with reference to Figure 7, providing more detailed analyses and interpretations. This includes an expanded explanation of the data patterns observed and their implications within the context of our study. Additionally, we have taken care to ensure that the methodological details are clearly presented earlier in the paper, making it easier for readers to understand the experimental setup and approach before delving into the results. These revisions aim to provide a more structured and comprehensible presentation of our research findings and methodologies, aligning with the standard format of research papers in our field.

6. The conclusion should include the extend of the future application or other.

Response: In response to your valuable feedback, we have refined our conclusion better to encapsulate the future application scope of our research. Specifically, we have expanded on the potential for fully blind localization, emphasizing the technical innovation of developing a system that operates effectively without prior layout-specific training. This advancement has significant implications for versatile deployment in dynamic environments. Furthermore, we have intensified our focus on the challenges and solutions related to multipath issues in backscatter communication systems. Our proposed methodology involves a comprehensive evaluation of environmental changes, such as the addition or removal of reflective surfaces, and their impact on signal integrity. We are exploring sophisticated pre-processing techniques within our machine learning algorithm to identify and adaptively respond to multipath components. This approach not only enhances the accuracy of our system but also opens avenues for broader applications in complex, real-world scenarios, thereby significantly contributing to the evolution of IoT networks and 6G technologies.

The newly added text is updated (in boldface) on page 12 of the revised manuscript.

7. The title why used potential? It may be improved with a accurately title.

Response: Thank you for your valuable feedback on the title of our paper. Upon reflection, we agree that the original title, "A Potential for Future AI-enabled Localisation with Energy Harvesting in Backscatter Communications for IoT," may inadvertently imply a speculative aspect of our research. To more accurately represent the concrete advancements and applications detailed in our study, we have revised the title to:

"Next-Generation IoT: Harnessing AI for Enhanced Localization and Energy Harvesting in Backscatter Communications."

We believe this new title confidently reflects the significant contributions of our work in advancing IoT technology. It directly conveys the integration and practical application of AI and energy harvesting in backscatter communications, emphasizing the innovative and forward-looking nature of our research.

Reviewer 3 Report

Comments and Suggestions for Authors

This paper explores using backscatter communication for indoor localization in future 6G networks. It provides an overview of backscatter communication and how it can enable low-power devices like IoT sensors in 6G networks. Backscatter allows devices to reflect and modulate existing RF signals instead of transmitting their own, enabling very low power operation. The authors build an RFID-based testbed with 8 tagged chairs to evaluate backscatter localization in a real indoor environment. They analyze the RF signal propagation and strength at different locations. The key innovations are using backscatter for joint indoor localization and energy harvesting, developing an ML model for localization, and demonstrating a complete system in a real environment. This explores new applications of backscatter in 6G and IoT.

The article has the following shortcomings:

1.The machine learning model may be overfitted to the specific testbed layout and environment. Testing on multiple setups would be needed to evaluate true generalization ability.

2. Only 8 tagged locations were used. Scaling up to many more tags in a larger space would reveal if the techniques work for more complex environments.

3. The localization accuracy degrades with distance from the reader antennas. Better techniques to compensate for signal attenuation over distance could improve performance.

4.The energy harvesting analysis is quite basic - more thorough modeling and measurements of harvested power would strengthen this part.

 

5. There is no comparison to other wireless localization methods like UWB or Bluetooth to benchmark performance. Backscatter needs to show advantages over existing technologies.

 

Author Response

Comment: This paper explores using backscatter communication for indoor localization in future 6G networks. It provides an overview of backscatter communication and how it can enable low-power devices like IoT sensors in 6G networks. Backscatter allows devices to reflect and modulate existing RF signals instead of transmitting their own, enabling very low power operation. The authors build an RFID-based testbed with 8 tagged chairs to evaluate backscatter localization in a real indoor environment. They analyze the RF signal propagation and strength at different locations. The key innovations are using backscatter for joint indoor localization and energy harvesting, developing an ML model for localization, and demonstrating a complete system in a real environment. This explores new applications of backscatter in 6G and IoT.

Response: Thank you for reviewing our manuscript. Your valuable feedback is highly appreciated. It indeed helped us in improving the quality of our paper.

Comment: The machine learning model may be overfitted to the specific testbed layout and environment. Testing on multiple setups would be needed to evaluate true generalization ability.

Response: Thank you for bringing up the concern about our machine-learning model potentially being overfitted to our specific testbed layout. We took various measures to address this, such as incorporating a diverse dataset covering different environmental scenarios. Furthermore, the regression plot for training, validation, and testing data also helps us with the machine-learning model to learn the patterns for the data effectively. However, we appreciate your suggestion of testing on multiple setups to evaluate the model's generalisation ability better. Therefore, we plan to extend our testing to different environments and collaborate with other research groups to assess our model's adaptability in various settings. This future work will help us understand the model's robustness and applicability in diverse conditions.

Comment:  Only 8 tagged locations were used. Scaling up to many more tags in a larger space would reveal if the techniques work for more complex environments.

Response: Thank you for your feedback regarding the scale of our experiment, particularly in terms of the number of tagged locations used in our study. We acknowledge that our initial setup involved only eight tagged locations and was intended to provide a proof of concept. However, we understand the importance of scaling up our experimental setup to test our techniques in more complex and varied environments rigorously. To address this, we are currently in the process of significantly increasing the number of tags and extending the physical space of our test environment. We aim to conduct more comprehensive experiments that will help us to improve our techniques and advance our research. We appreciate your suggestion, as it aligns with our current efforts and the future trajectory of our study. We will keep you updated on the progress of our expanded experiments and look forward to sharing the results in future work. 3. The localisation accuracy degrades with distance from the reader antennas. Better techniques to compensate for signal attenuation over distance could improve performance.

Comment: The energy harvesting analysis is quite basic - more thorough modelling and measurements of harvested power would strengthen this part.

 Response: Thank you for your valuable feedback on our study's energy harvesting aspect. We appreciate your observation that our current analysis in this area is basic, and we acknowledge the importance of a more thorough and detailed approach to modelling and measuring harvested power.  As per our earlier discussion, we are expanding our existing experimental setup. This involves better implementing more sophisticated modelling techniques to capture the nuances of energy harvesting in RFID systems. We also plan to conduct a series of comprehensive measurements covering a broader range of operational conditions and utilise more sensitive and precise instrumentation to quantify the harvested power accurately.  Once again, we thank you for your feedback and suggestions, which have been instrumental in helping us improve our study.

Comment: There is no comparison to other wireless localization methods like UWB or Bluetooth to benchmark performance. Backscatter needs to show advantages over existing technologies.

Response: Thank you for suggesting that we include a comparison of our backscatter localisation method with other wireless localisation technologies such as Ultra-Wideband (UWB) or Bluetooth. We acknowledge the significance of benchmarking performance against existing technologies to demonstrate new methods' relative strengths and potential advantages. It is important to note that the main objective of this paper is to develop and validate the RFID-based backscatter localisation technique, specifically in the context of energy harvesting in sensing networks. The paper's scope is intentionally focused on demonstrating the feasibility and effectiveness of this technology in controlled environments. Therefore, while comparing this technique to other wireless localisation methods would undoubtedly be valuable, it is beyond the scope of this research.

Reviewer 4 Report

Comments and Suggestions for Authors

- Page 5, line 220: In what frequency band are the acceptable levels?

- How far does your solution allow you to power devices?

- How was the heat map measured?

- Page 7, line 273: What was the amount of energy harvested?

- What was the algorithm implemented in?

- How did the results of algorithms KNN and Random Forest differ?

- How is the accuracy of the results related to the number of antennas?

Author Response

Comment: Page 5, line 220: In what frequency band are the acceptable levels?

Response: Thank you for pointing out the need for specificity regarding the frequency band. The frequency range of 550 MHz to 2.5 GHz.

 Comment:  How far does your solution allow you to power devices?

Response: Regarding your question about the range of our solution for powering devices: Our system, primarily focused on localisation, also enables energy harvesting. The exact range for powering devices would depend on the efficiency of the energy harvesting mechanism and the power requirements of the devices. However, our primary aim was to enhance the accuracy of localisation within a controlled environment​​.

 Comment:  How was the heat map measured?

Response: The heat map was measured through a systematic process of data collection in our testbed setup. We used RSSI and phase angle parameters to create a fingerprint of the area, which was then used to train our machine learning models. This method allowed us to generate a detailed representation of the signal strength and quality across different points in the testbed​​.

 Comment: Page 7, line 273: What was the amount of energy harvested?

Response: In response to your query about the amount of energy harvested: Our paper primarily discusses the advancements in localisation using backscatter communication. While we mention the potential of integrating energy harvesting, the specific quantification of energy harvested was not the focus of this study and will be explored in more depth​​ in future work.

 Comment:  What was the algorithm implemented in?

Response: The algorithms, K-Nearest Neighbours (KNN) and Random Forest were implemented in a supervised machine learning training environment (python). We used RSSI and phase angle data collected from the testbed setup to train these algorithms. The choice of these algorithms was based on their suitability for classification tasks in our localisation application​​.

 Comment:  How did the results of algorithms KNN and Random Forest differ?

Response: Concerning the differences between KNN and Random Forest results: We found Random Forest to be more advantageous mainly due to its scalability and ability to handle live data more effectively. KNN, while simpler and more interpretable, did not scale as well. The performance advantage of Random Forest was particularly evident in environments with a high volume of data and the need for up-to-date results​​.

 Comment:  How is the accuracy of the results related to the number of antennas?

Response: The accuracy of our results showed a direct correlation with the number of antennas used. Using two antennas yielded an accuracy of 82% with the Random Forest algorithm, which improved to 93% upon adding a third antenna. This indicates that additional antennas significantly enhance the system's localisation accuracy, although they may not be feasible for all applications due to cost considerations​​.

 

Reviewer 5 Report

Comments and Suggestions for Authors

Dear Authors,

The aim of this paper is analyzing the current state of backscatter communication, and it’s potential in future 6G networks, and we have looked at the state of the art of RF localization.

You have created a system that is able to localize down to set locations even when the environmental changes continuously affect the signal strength.

The strengths of the paper are the analysis of the limitations of the BSC technology and its potential impact on 6G networks, the presentation of the system model, including the arrangement of BSC tags for optimal performance and the positioning of the RF antenna. Another strength is the examination of the implementation of the approach, providing a comprehensive overview of the techniques and methods used. The authors also analyze how the approach performed and the impact of changing parameters. The authors provide a conclusion of the authors´ findings and propose potential applications for the paper. 

This is a great paper, with 44 references. The literature review is well organized, and all the cited references are relevant to the research. The introduction provides sufficient background and include all relevant references.

The English language is fine. No issues detected.

However, there are a few aspects to be fixed before its publication:

- The Author Contributions are described at all. There is not any information about who contributed to what.

- Some of the figures should be explained better.

Author Response

Comment: The aim of this paper is analyzing the current state of backscatter communication, and it’s potential in future 6G networks, and we have looked at the state of the art of RF localization.

Response: We appreciate your feedback on the aim of our paper. In light of your comments, we have updated the abstract to more clearly articulate our analysis of the current state of backscatter communication and its potential in future 6G networks, with a specific focus on RF localization. This revision better highlights our contribution to the field and aligns with the objectives of our study.

Comment: You have created a system that is able to localize down to set locations even when the environmental changes continuously affect the signal strength.

Response: Thank you for recognizing our work on creating a dynamic localization system. We agree with your observation and have further elaborated on this aspect in the results section. More details, especially regarding the accuracy graphs under varying environmental conditions, have been added to demonstrate the robustness and adaptability of our system.

Comment: The strengths of the paper are the analysis of the limitations of the BSC technology and its potential impact on 6G networks, the presentation of the system model, including the arrangement of BSC tags for optimal performance and the positioning of the RF antenna. Another strength is the examination of the implementation of the approach, providing a comprehensive overview of the techniques and methods used. The authors also analyze how the approach performed and the impact of changing parameters. The authors provide a conclusion of the authors´ findings and propose potential applications for the paper. 

Response: We are grateful for your positive assessment of the strengths of our paper. In response, we have made sure to further refine these sections, particularly emphasizing our analysis of the limitations of Backscatter Communication (BSC) technology in 6G networks, our system model presentation, and the implementation details. Your feedback has been instrumental in enhancing the clarity and depth of our discussion.

Comment: This is a great paper, with 44 references. The literature review is well organized, and all the cited references are relevant to the research. The introduction provides sufficient background and include all relevant references.

Response: Thank you for your encouraging words about our paper and its literature review. We are pleased to know that our efforts in organizing the literature and providing a comprehensive introduction have been well received. Your acknowledgement serves as a valuable affirmation of the rigour and relevance of our work.

Comment: The English language is fine. No issues detected.

Response: We are glad to hear that the English language used in the paper meets the necessary standards. We have made every effort to ensure clarity and precision in our communication.

Comment: However, there are a few aspects to be fixed before its publication:

Response: We appreciate your note on the aspects that require attention. Following your guidance, we have addressed these points in the revised manuscript to ensure that it meets the publication standards.

Comment: - The Author Contributions are described at all. There is not any information about who contributed to what.

Response: We acknowledge the oversight in not detailing the Author Contributions. The revised manuscript now includes a clear section that outlines the specific contributions of each author. This addition provides transparency and acknowledges the collaborative efforts of our research team.

Comment: - Some of the figures should be explained better.

Response: In response to your comment, we have revised the descriptions of the figures in the paper for better clarity. Each figure now includes a more detailed explanation, ensuring that they effectively complement and elucidate the text. This revision enhances the overall comprehensibility of our research findings. The updated text in our revised manuscript is in boldface text.

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

According to the comments of the reviewer, the authors of manuscript have made detailed revisions to this manuscript, making the manuscript’s content more abundant and reasonable, and the language more fluent. The formula format meets the requirements of Electronics. Therefore, the reviewer believes that the manuscript can be acceptable.

Reviewer 2 Report

Comments and Suggestions for Authors

The paper was revised well.

Reviewer 3 Report

Comments and Suggestions for Authors

This research have several potential implications and applications for the future of IoT networks, including:

1. Improved Localization: The use of backscatter communication and machine learning frameworks can enable more precise and reliable localization of IoT devices in dynamic real-world environments.

2.  Energy Harvesting: Backscatter communication systems can be used to harvest ambient RF energy, which can be used to power low-energy IoT devices perpetually.

3. Cost-Effective Solutions: Backscatter communication systems are cost-effective compared to battery-powered sensors, making them an ideal solution for low-power and low-cost IoT devices.

4. Coexistence with Existing Technologies: Backscatter communication can coexist with existing wireless technologies, making it an ideal solution for dense urban environments where spectrum availability is a major challenge.

5. Support for 6G Networks: Backscatter communication is expected to play a critical role in strengthening the next generation of wireless communication, supporting the ultra-high-speed, ultra-low-latency, and ultra-high-reliability communication requirements of 6G networks.

In the future, I hope the authors can provide more scenario applications.