kumoh national institute of technology
Networked Systems Lab.

Review Comment

NSL > Works@NSL> About Review> Review Comment
Rizki Rivai Ginanjar, and Dong-Seong Kim, "ELM-Based Real-Time Node Localization: UAV Approach", IEEE Wireless Communication and Networking Conference (WCNC) 2019, April 15th - 18th, 2019, Marrakech, Maroko (R)
By :
Date : 2019-02-07
Views : 194

Dear Mr. Ginanjar,

We regret to inform you that your paper #1570495053, 'ELM-Based Real-Time Node Localization: UAV Approach', has not been accepted for presentation at the 2019 IEEE Wireless Communications and Networking Conference (WCNC).

The conference received a very high number of submissions and was thus very competitive. Each paper was carefully peer-reviewed by at least three reviewers and/or TPC members. The reviews for your paper are given below and can also be found at https://edas.info/showPaper.php?m=1570495053.

We would like to thank you for submitting your manuscript to the conference. However, due to presentation slot limitations, many good and/or interesting papers could not be included in the program.

Understanding that the result is disappointing to you, we hope nonetheless that you will consider submitting again to future editions of WCNC.

Sincerely,
TPC Track Chairs
IEEE WCNC 2019

========================

======= TPC Review 1 =======

> *** TPC Recommendation: Given the reviews, what is your recommendation for this paper.
Weak reject (2)

> *** TPC Summary: Please give a justification for your recommendation, especially if the review scores vary widely or if your recommendation differs significantly from those of the reviewers.

It is not explicitly explained how the proposed neural network (NN) algorithm reduces the complexity overhead (as it is mentioned in the introduction)

======= Review 2 =======

> *** Strong Aspects: Comments to the author: What are the strong aspects of the paper?

The paper addresses the problem of WSN node localization using UAV and neural networks.

> *** Weak Aspects: Comments to the author: What are the weak aspects of the paper?

It is not clear what is new in the paper, given that the ELM scheme was already used before for node localization. There are many inconsistencies with the material presented in the paper. The results of the scheme show large error for this and the other localization schemes. Performance evaluated through simple simulation and lack of experimental evaluation is insufficient for practical verification.

> *** Recommended Changes: Recommended changes. Please indicate any changes that should be made to the paper if accepted.

* As the UAVs are small, the RSSI difference between its two antennas will be small too, especially as compared to the propagation impairments, so the two-antenna system may not work well in practice.

* It is not clear how the UAVs know how to fly along the blocks. Do the UAVs rely on GPS?

* The simulations assume in-building and line-of-sight environment. In such an environment there is no problem relying on fixed anchor points for localization. So, this is not a satisfactory evaluation setup.

* In the in-building and line-of-sight environment there will typically be a lot of multipath fading, but the evaluation does not take this into consideration.

* Presumably the results will be different with different training environment. So training is environment dependent and there is already the need for fixed nodes for training. Why those nodes cannot serve as anchors for localization also?

* The results show very large error for all the three schemes, suggesting that this may not be a practical approach.

* The paper suggests that the error could be reduced with multiple UAVs, but no results are presented.

* The paper should be reviewed for English presentation style.

> *** Relevance and Timeliness: Rate the importance and timeliness of the topic addressed in the paper within its area of research.
Acceptable. (3)

> *** Technical Content and Scientific Rigour: Rate the technical content of the paper (e.g.: completeness of the analysis or simulation study, thoroughness of the treatise, accuracy of the models, etc.), its soundness and scientific rigour.
Questionable work with severe flaws. (1)

> *** Novelty and Originality: Rate the novelty and originality of the ideas or results presented in the paper.
Minor variations on a well investigated subject. (2)

> *** Quality of Presentation: Rate the paper organization, the quality of text, English, and figures and the completeness and accuracy of references.
Readable, but revision is needed in some parts. (3)

======= Review 3 =======

> *** Strong Aspects: Comments to the author: What are the strong aspects of the paper?

Figures are clear for readers to understand the problem buildup.

> *** Weak Aspects: Comments to the author: What are the weak aspects of the paper?

1. Writing needs to be improved.
2. Problem formulation is not persuasive. I hardly agree that this method is scalable. When the area increases, you need more UAVs to cover. Also, since the RSSI pattern can be learned, I think only one UAV is required for locating a row/column of blocks.
3. Lack of background introduction. Why do you choose UAV to do the localization? What is the goal of your work? Energy? Accuracy?
4. The evaluation is weak. You only have one table and two figures to show some evaluation results, while in the two figures, your method doesn't show any superiority.

> *** Recommended Changes: Recommended changes. Please indicate any changes that should be made to the paper if accepted.

1. Please reforge the writing.
2. Do not use notations like Rx, Rss. Use R_x, R_{ss} instead.
3. Please describe the background. Maybe some motivations for this work.
4. Please evaluate with some more details.

> *** Relevance and Timeliness: Rate the importance and timeliness of the topic addressed in the paper within its area of research.
Acceptable. (3)

> *** Technical Content and Scientific Rigour: Rate the technical content of the paper (e.g.: completeness of the analysis or simulation study, thoroughness of the treatise, accuracy of the models, etc.), its soundness and scientific rigour.
Marginal work and simple contribution. Some flaws. (2)

> *** Novelty and Originality: Rate the novelty and originality of the ideas or results presented in the paper.
Minor variations on a well investigated subject. (2)

> *** Quality of Presentation: Rate the paper organization, the quality of text, English, and figures and the completeness and accuracy of references.
Unacceptable. (1)

======= Review 4 =======

> *** Strong Aspects: Comments to the author: What are the strong aspects of the paper?

Try to use ELM in the positioning system, aided by UAVs as anchors.

> *** Weak Aspects: Comments to the author: What are the weak aspects of the paper?

Although ELM outperforms the SVM (I doubt this heavily), the localization is still ver poor. We know that RSSI is a vulnerable parameters in wireless transmission, which limits the application in localization.

> *** Recommended Changes: Recommended changes. Please indicate any changes that should be made to the paper if accepted.

The authors are encouraged to improve the positioning accuracy by choose other parameters such as the trajectory of anchors.

> *** Relevance and Timeliness: Rate the importance and timeliness of the topic addressed in the paper within its area of research.
Good. (4)

> *** Technical Content and Scientific Rigour: Rate the technical content of the paper (e.g.: completeness of the analysis or simulation study, thoroughness of the treatise, accuracy of the models, etc.), its soundness and scientific rigour.
Solid work of notable importance. (4)

> *** Novelty and Originality: Rate the novelty and originality of the ideas or results presented in the paper.
Minor variations on a well investigated subject. (2)

> *** Quality of Presentation: Rate the paper organization, the quality of text, English, and figures and the completeness and accuracy of references.
Well written. (4)

======= Review 5 =======

> *** Strong Aspects: Comments to the author: What are the strong aspects of the paper?

A single hidden layer feedforward neural network (SLFN) learning algorithm is proposed for real-time localization technique, based on RSSI values of the beacon signals from mobile anchor nodes. Extreme learning machine (ELM) is used as the learning method to train the SLFN. To improve the localization efficiency unmanned aerial vehicle (UAV) is used in the WSN because of its flexibility of movement to cover whole area by moving from one coverage area to other. The proposed algorithm uses a single hidden-layer in the neural network to minimize training and testing time compared to other related algorithms. Also, there is no on-the-ground anchor node, offering better scalability compared to other traditional methods, as there is no need to reconfigure the previously deployed anchor nodes if the network size is increasing. Simulation results show that the proposed ELM algorithm outperforms other algorithms (the backpropagation and support vector machine (SVM) learning method) i!
n terms of network training and testing time.

> *** Weak Aspects: Comments to the author: What are the weak aspects of the paper?

It is not explicitly explained how the proposed neural network (NN) algorithm reduces the complexity overhead (as it is mentioned in the introduction), in comparison to previous works focusing on localization techniques in the WSN system, utilizing UAVs to estimate a node position, which they do not use NNs. Also, it is not explicitly explained what is the contribution of the proposed solution, in comparison to previous works that use SLFN and ELM on the node localization process. For example, is it the first time that a UAV-based solution is proposed that uses SLFN and ELM?

> *** Recommended Changes: Recommended changes. Please indicate any changes that should be made to the paper if accepted.

It is recommended to explain how the proposed neural network (NN) algorithm reduces the complexity overhead in comparison to previous works focusing on localization techniques in the WSN system, utilizing UAVs to estimate a node position, which they do not use NNs; or not to mention complexity minimization, but another possible advantage. Also, it is recommended to explicitly explain the contribution of the proposed solution, in comparison to previous works that use SLFN and ELM on the node localization process.

> *** Relevance and Timeliness: Rate the importance and timeliness of the topic addressed in the paper within its area of research.
Good. (4)

> *** Technical Content and Scientific Rigour: Rate the technical content of the paper (e.g.: completeness of the analysis or simulation study, thoroughness of the treatise, accuracy of the models, etc.), its soundness and scientific rigour.
Valid work but limited contribution. (3)

> *** Novelty and Originality: Rate the novelty and originality of the ideas or results presented in the paper.
Some interesting ideas and results on a subject well investigated. (3)

> *** Quality of Presentation: Rate the paper organization, the quality of text, English, and figures and the completeness and accuracy of references.
Well written. (4)