A Long-Range Internet of Things-Based Advanced Vehicle Pollution Monitoring System with Node Authentication and Blockchain

Round 1

Reviewer 1 Report

The revised manuscript still has two key weaknesses in my opinion:

First, very little information is provided about the sensors used to monitor vehicle pollution (i.e., their manufacturer, their ranges of validity, any field or lab calibrations conducted for the sensors and their performance during characterization). As it stands, I have no frame of reference with which to judge how well these sensors can measure the actual pollutant concentrations or emissions from the vehicles. What remains, therefore, is for me to assume that these measures are valid, or can be made valid with appropriate calibration, and focus on the LoRA communication and blockchain verification schemes proposed here as being the main contribution of the paper.

However, second, there are no objective performance metrics or benchmarks with which I can evaluate the value of the LoRA network and/or blockchain authentication over any other existing or proposed system architecture. In order to do this, the expected benefits of these new features must be clearly stated and quantified, experiments must be designed to test the capabilities of these new features, and an evaluation must be made of whether the proposed architecture does or does not meet expected performance criteria. The only presented results indicate that a single sensor node was able to collect some amount of data; no information is provided about the quality of these data compared to an independent measurement, or how well the technical characteristics of the node (e.g., communication range, battery life, data throughput, integrity of the data) matched with expected performance goals.

In summary, the paper only serves to demonstrate that the proposed network architecture is feasible and has been implemented in a small scale. It does not quantify or demonstrate what benefits this new architecture has over other architectures (for example, by testing the maximum communication ranges between nodes to verify that the LoRA system is performing as expected, or demonstrating how the blockchain system can prevent erroneous or malicious logging of incorrect air quality data), nor does it even demonstrate that an individual sensor node is taking accurate measurement by comparing it to a known standard. In my opinion, while potentially of use to guide future technology development, the paper does not present any testable hypothesis related to the proposed architecture, which makes it impossible for me to evaluate how useful such an architecture would be in addressing the problem of monitoring vehicular pollution at a large scale.

Author Response

Dear Reviewer,

Thank you for spending time on our manuscript and providing us an opportunity for submitting the revised draft of our manuscript. We would like to acknowledge you for your comments and suggestions that helped us improve the manuscript. We have considered the comments with care and addressed them with the best of our efforts and knowledge. Each comment is point by point answered in attachment for your consideration.

Author Response File: Author Response.pdf

Reviewer 2 Report

This revision is improved.

Author Response

Dear Reviewer,

Thank you for spending time on our manuscript and helping us in its improvement.

Reviewer 3 Report

The authors correct all of my remarks, so I accept the article in present form.

Author Response

Dear Reviewer,

Thank you for spending time on our manuscript and helping us in its improvement.

Reviewer 4 Report

Well structured and well written manuscript. I Recommend it for publication. 

Author Response

Dear Reviewer,

Thank you for spending time on our manuscript and helping us in improving its quality by your valuable suggestions. 

Round 2

Reviewer 1 Report

Thank you for providing information about the types of sensors being used. Please note that low-cost air quality sensors, such as the ones you have mentioned, require extensive calibration and verification of their data in different deployment environments. If you do not perform such verification yourselves, I would suggest citing other work which uses the same types of sensors in similar environments to yours and reports their performance. Also note that these sensors are designed for ambient atmospheric monitoring. As such, they may not be suitable for monitoring vehicle emissions directly, which seems to be the intended use in this application. I am unaware of any work specifically relating to this application specifically; I would suggest citing any relevant work you can find for the use of these particular sensors for vehicle emissions monitoring. For example, this articles looks at the use of low-cost air quality sensors for estimating fleet-level vehicle emissions:

 

https://doi.org/10.1016/j.trd.2020.102635

 

However, measurement of emissions for individual vehicles using low-cost sensors mounted on the vehicles themselves was not attempted in that work. These limitations (lack of verification of the data against a standard method for vehicle emissions measurement and potential unsuitability of these particular low-cost sensors for this purpose) should be mentioned in the conclusions and left as a topic for future work.

 

With regards to the technical details of the LoRA system you provide, these are useful, but I am still missing some important context related to these results. In particular, a performance target or targets are needed for the metrics you present. In other words, what levels of sensitivity, data rate, etc., would be required for collecting air quality data from the proposes system for a realizing number of devices mounted on vehicles. The data presented seems to be what is required to answer this type of question, but the answer itself is lacking.

 

Finally, it is still unclear to me what additional capabilities or benefits the integration of a blockchain system into the proposed framework provides. As the use of such a system is a key difference between this work and others, as noted in table 15, and so I would expect verification and validation of this capability to be a major topic of discussion. For example, the performance of the system (in terms of data rates and security) could be compared with a similar system implemented without the blockchain aspect. In other words, what are the benefits and drawbacks (assessed in as quantitative a way a possible) to incorporating blockchain for verification and validation within this architecture, as opposed to another verification system?

Author Response

Dear Reviewer,

Thank you for spending time on our manuscript and providing us an opportunity for submitting the revised draft of our manuscript. We would like to acknowledge you for your comments and suggestions that helped us improve the manuscript. We have considered the comments with care and addressed them with the best of our efforts and knowledge. Each comment is point by point answered in attachment for your consideration.

Author Response File: Author Response.pdf