LoRaWAN Path Loss Measurements in an Urban Scenario including Environmental Effects

Round 1

Reviewer 1 Report

In this paper a dataset is presented for LoRaWAN path loss modeling. LoRawan path loss modeling is very important for many IoT applications. So this dataset is very useful and can be helpful for many applications like design a LoRaWAN network or improvement the RSSI based localization approaches. My comment about this dataset is as follow:

1-      Please add some explanations about the position of node antenna. Do you check that antenna of node EN1, EN2 and En3 be located 90 degrees and vertical? What about the En4 antenna?

2-      Please add some explanations about the frequency of data based on the different SF. It seems the path loss model depends to the SF. So it is better to group the dataset for each SF separately.

3-      Based on table 3, why didn’t use SF 11 and 12? These SFs are very important for long range.

4-      Please add some explanations about the packet loss of each node.

Author Response

Please see the attachment

Author Response File: Author Response.pdf

Reviewer 2 Report

The structure of the document lacks a logical order. 

The quality of the figures needs to be improved. It is recommended to use Matlab and the command print -dpdf -r800 figure1 to achieve an acceptable rate. 

Using Overleaf as a word processor is recommended to ensure the document's quality.

The references are recommended to use a range between 2023 and 2019. In the section on related works, make a summary table with what other authors have proposed to solve the problem and highlight the innovation of this work.

 

On the other hand, it is suggested that the authors order the document.

1.- Introduction (Generalities of the problem - 15 references between 2023-2019), 2.- Related Works (Specific works with punctual solutions, summary table of state of the art of other proposals against the current work - 5 references between 2023-2019), 3.- Problem Formulation and Methodology (Table of variables of the mathematical model, mathematical model, pseudocode, methodology flowchart, Big Oh Notation of the algorithm), 4. - Analysis of Results (Metrics performed in Matlab in PDF format or directly from Overleaf), 5.- conclusions (Direct relation between the objective stated in the abstract vs. the metrics found), 6.-References (All documents with DOI from Sciencedirect, MDPI, Wiley, PLOS, Taylor & Francis, Springer, Hindawi, IEEE Xplore [transactions, magazines, journals]). 

It is recommended to evaluate the routing with the wireless technology presented in this document in a geo-referenced scenario. The innovation of your work should be relevant to publications between 2023 and 2019. The structure of the paper requires modification. They should increase the proposal's performance metrics considering the sensors' expansion and scalability. The problem of significant relevance is the loss pattern for an investment that evaluates the network's evolution and scalability and the cost of wireless technology. It is suggested to add an optimization model. It is recommended to use Openstreetmap and Matlab to evaluate an urban, rural, and suburban area.

 

Author Response

Please see the attachment

Author Response File: Author Response.pdf

Reviewer 3 Report

The authors conduct a long-term testing campaign of LoRaWAN deployed in a mountainous location. They used 4 nodes and one gateway to gather data for 6 months with nearly 1 million transmissions. The author gathered very detailed parameters that could be relevant to modelling of LoRaWAN transmissions. The authors demonstrate the use-case for their dataset with two example models.

The authors have produced a great dataset that can be helpful in LoRaWAN modelling. However, I can't help but wonder how much the ideosyncratic structure of the city is impacting the transmission. It is not clear how these four nodes can be transfered to a new geometrically different location. It would be great to have more data points from different locations across the city. I know that this may be out of scope but it limits the applicability for modelling in other places.

For a dataset, I am also missing the data set descriptives. On which SF was the data send mostly (especially with ADR? Did ADR bias the SF selection (e.g. only SF10 or SF7)? Which frequencies where used by the real data, is it equally distributed? We have 192 combinations according to the authors, have they been equally distributed? Do we have any biases? Why? Is the data temporally different distributed across the 6 months? Is the SF distributed different for the 4 devices? I would also like to see the output measure distributions including for the novel parameter of energy. Please bring in a clear discription of the distribution of your data.

Another thing that could be improved are the provided metrics. ESP has been proposed as an alternative to RSSI and SNR (see [1] for definition and [2] for application). I would implore the authors to add ESP to their analysis especially since it can be calculated post hoc. I think that to complement these analysis it could be very useful. There has been major efforts to model not only explanatory models but also exporatory models. Especially Transformers, but also LSTM models have been proven to be effective [2]. It would be great to see these models applied to the data. Especially a variance analysis showing how each factor is impacting the predictive models could show how the these newly discussed factors matter in a more hands-on way.

 

I think it is hard to say whether more data should be collected and how difficult that is but I would ask the authors to improve the data description and also use some predictive models to show the usefulness of the data set.

 

Refs:

[1] Abdelghany, A., Uguen, B., Moy, C., & Lemur, D. (2021, June). On Superior Reliability of Effective Signal Power versus RSSI in LoRaWAN. In 2021 28th International Conference on Telecommunications (ICT) (pp. 1-5). IEEE.

[2] Grübel, J., Thrash, T., Aguilar, L., Gath-Morad, M., Hélal, D., Sumner, R. W., ... & Schinazi, V. R. (2022). Dense Indoor Sensor Networks: Towards passively sensing human presence with LORAWAN. Pervasive and Mobile Computing, 84, 101640.

Author Response

Please see the attachment

Author Response File: Author Response.pdf

Round 2

Reviewer 2 Report

Table 1 and Table 2 should be merged.

 

It is recommended that a geo-referenced scenario for routing be provided in the future. It is recommended to use Openstreetmap and Matlab to generate an algorithm in a specific area and identify the scalability of the users.

 

It is recommended that the references have DOI. Those references that are not scientists should be replaced.

Author Response

Response letter for Manuscript Data-2057962

LoRaWAN Path Loss Measurements in an Urban Scenario Including Environmental Effects

Mauricio González-Palacio, Diana Tobón-Vallejo, Lina M. Sepúlveda Cano, Santiago Rúa, Giovanni Pau, and Long Bao Le

We want to thank the editor and the reviewers for your valuable comments and suggestions, which have significantly helped us improve our paper. We have considered all the points raised by the reviewers to revise our paper. 

Response to Reviewer 2

"It is recommended that a geo-referenced scenario for routing be provided in the future. It is recommended to use Openstreetmap and Matlab to generate an algorithm in a specific area and identify the scalability of the users."

We thank your suggestion and will add it to our future work since it is a relevant topic we have not tackled yet. 

"It is recommended that the references have DOI. Those references that are not scientists should be replaced."

We would like to address this comment as follows. We added the DOIs to all the references as recommended by the reviewer.

 

Reviewer 3 Report

The authors have taken great care to expand on the reviewer comments. Additional, measures, descriptives and figures were provided in line ith the reviewer's request.

Author Response

Response letter for Manuscript Data-2057962

LoRaWAN Path Loss Measurements in an Urban Scenario Including Environmental Effects

Mauricio González-Palacio, Diana Tobón-Vallejo, Lina M. Sepúlveda Cano, Santiago Rúa, Giovanni Pau, and Long Bao Le

We want to thank the editor and the reviewers for your valuable comments and suggestions, which have significantly helped us improve our paper. We have considered all the points raised by the reviewers to revise our paper. In the following, we provide our responses to the reviewers' comments. 

Response to Reviewer 3

"The authors have taken great care to expand on the reviewer's comments. Additional measures, descriptions, and figures were provided per the reviewer's request."

Thank you for your comments. Our paper improved after adding and fixing all the points raised by you.