Performance Enhancement of Optimized Link State Routing Protocol by Parameter Configuration for UANET
Round 1
Reviewer 1 Report
Generally well written article with clear understandable language, it would be interesting for the readers. The main contribution of the article includes first, the new Enhanced OLSR routing protocol for UANET and, second, experiments and comparison for three performance indicators for 5 routing protocols including E-OLSR. However there is no proper description of the proposed E-OLSR, there is only fig. 1 with some E-OLSR parameters and fig. 2 with basic packet layout (and it is actually OLSR layout, not the E-OLSR one). It would be reasonable to present some more or less formal scheme of the protocol, emphasizing its novelty in comparison with OLSR. Moreover for a new protocol it would be reasonable to show at least some elements of the formal proof that the protocol is formally correct (it is desirable to present how the authors could check the correctness of this protocol).
As for the simulation modeling of the used five routing protocols in OPNET tool, there is almost no due information, how exactly they were modeled there. Especially for E-OLSR (it is stated, for instance in lines 296-297, that it is really proposed in this article). And if for the rest 4 routing protocols the authors used already available models constructed by anyone before, this fact should be mentioned as well and due references to them would be given.
The simulated scenarios for UAVs are described too briefly, it is a lack of information how and when they take off, land and move, is it a chaotic movement, organized on a base of probability-based determined manner or something else (fig. 3 seems not to be over-informative, while fig. 5 and 6 captions exposes some basic data on UAV speeds, altitudes and data rates). Besides it'd be interesting to learn the models of UAVs the authors are considering when doing the simulation.
The experimental results are seen as quite well, but their analysis is straightforward and superficial (basically, the analysis comes down to which metrics are better for each considered routing protocol).
It is reasonable to highlight more explicitly why the performed simulation modeling is specific exactly to UANET (and not for other close types of networks analyzed/mentioned in the first part of the articled). In lines 102-103 it is mentioned about some principal differences between UANET and MANET, but more information on it would be given.
In introduction it is justified, why the performance is really needed for such UANET routing protocols. But for greater reliability it could be proved also reasoning from any particular examples of UANETs and their business requirements/limitations in a specific application field.
For the uniformity for Data Drop Rate it'd be better to present a calculation formula (it may be too simple as a quotient of the delivered packets to all the ones, but nevertheless it should be exposed).
In lines 57-58 there are abbreviations that are deciphered in a few pages later only.
In fig. 2 the font inside the boxes could be a bit larger.
In fig. 4 rsvp is poorly readable due to the arrows.
A misprint in line 224, it should be 30 m\s instead of 30 m\.
Summarizing, the authors made a great job, and I think after all the corrections to be done the article could be accepted.
Author Response
Dear Reviewers and Editors,
First of all, we (the authors) want to express our gratitude to the reviewers and editors for the thoughtful and helpful comments on an earlier draft of our manuscript, which is entitled “Compara-tive Evaluation of Enhanced Optimized Link State Routing with Topology-based Routing Protocols for UANET”. Your letter indicated that our manuscript (ID: drones-1510388) must be substantially revised to be considered for publication. We have made nearly all of the requested edits and feel as though this revised manuscript is vastly improved and ready for publication. What follows is a detailed list of the concerns and how we addressed them.
The response letter is attached in a file with through discussion. Please see the attachment.
Cordially,
The authors
Author Response File: 
Author Response.pdf
Reviewer 2 Report
In the state of art, the four routing protocols must be specified in terms of applicability tp the drone context, pros and cons, etc.. A final comparison table should help to sistemize the main characteristics of each of them. The main common features are to be identified, together with the differences.
The E-OLSR protocol is not described.
The scientific approach and ratio are not described.
The contribution of the authors is not clear. It must be specified, described. The differences with the other routing protocols must be described.
The simulation set up is not specified and justified, related to the drone context.
IP impairments and background traffic are not considered in the performed simulations. It provides a lack of credibility of the simulations. They are not realistics.
The influence of the different metrics is not treated, neither a final comparison of them is provided.
The shown results are simple, insufficient and not suitable for a scientific paper: they are a simple screeshot from the commercial OPNET simulation tool.
Author Response
Dear Reviewers and Editors,
First of all, we (the authors) want to express our gratitude to the reviewers and editors for the thoughtful and helpful comments on an earlier draft of our manuscript, which is entitled “Comparative Evaluation of Enhanced Optimized Link State Routing with Topology-based Routing Protocols for UANET”. Your letter indicated that our manuscript (ID: drones-1510388) must be substantially revised to be considered for publication. We have made nearly all of the requested edits and feel as though this revised manuscript is vastly improved and ready for publication. What follows is a detailed list of the concerns and how we addressed them.
The response letter is attached in a file with a thorough discussion. Please see the attachment.
Cordially,
The authors.
Author Response File: Author Response.pdf
Reviewer 3 Report
The paper is interesting and addresses a very timely topic. Please find below the main parts that need to be improved:
1) In my opinion, the initial part of the Introduction can be improved a bit more. Authors presented some of the most important application fields of UAV technologies, and clearly explained the major benefits related to their use compared to other systems. However, in order to provide a complete overview on UAVs, I think that also the possible raising threats associated to a malicious use of UAVs technologies deserve to be discussed. A prominent example is the use of small and cheap UAVs to perform illegal activities such as drug smuggling, privacy violation, and even terrorist attacks. In this respect, I would suggest adding some quite recent and interesting work on the topic:
- A. Schumann et al, "Drone vs. Bird Detection: Deep Learning Algorithms and Results from a Grand Challenge", Sensors, 2021
- S. Jamil et al, "Malicious UAV Detection Using Integrated Audio and Visual Features for Public Safety Applications", Sensors, 2020.
but authors are encouraged to add also other references they may be aware of.
2) Starting from line 66, authors discuss the three main category of routing protocols. An aspect that I think is too much understated is the importance of having accurate estimates of the UAVs trajectories over time. Indeed, an inaccurate knowledge of UAV positions negatively impacts onto most routing protocols, not only on the position-based ones, especially when multiple UAVs need to be coordinated. As correctly pointed out, the major difficulties reside in the very fast maneuvering capabilities of UAVs, which call for the need of advanced localization methods. In this respect, it would be useful to provide some pointer to the existing literature on UAV localization, which is necessary for any routing protocol.
3) A Table summarising the main benefits and drawbacks of the original OLSR, AODV, DSR, and GRP protocols discussed in Sec. 2 would be helpful to corroborate the comparison among them and highlight at the same time the improvements brought by the proposed enhanced OLSR (E-OLSR).
4) Please consider using some mathematical symbols instead of the extended names in eq. (1). It is very unusual to indicate the total number of bit sent and the total data sending time with their complete name. For instance, consider using N_b for the number of bits and T_tot for the total sending time. The same consideration applies to the throughput.
5) How many independent runs of the simulation have been performed to estimate the average throughputs reported in Fig. 5, the average delays in Fig. 6, and the average data dropped in Fig. 7?
6) If not strictly necessary, please limit the x-axis values to the range of values used to test the protocols in the simulation, namely 0-600 seconds. Legends can be easily moved in some other part of the graph.
Author Response
Dear Reviewers and Editors,
First of all, we (the authors) want to express our gratitude to the reviewers and editors for the thoughtful and helpful comments on an earlier draft of our manuscript, which is entitled “Comparative Evaluation of Enhanced Optimized Link State Routing with Topology-based Routing Protocols for UANET”. Your letter indicated that our manuscript (ID: drones-1510388) must be substantially revised to be considered for publication. We have made nearly all of the requested edits and feel as though this revised manuscript is vastly improved and ready for publication. What follows is a detailed list of the concerns and how we addressed them.
The response letter is attached in a file with a thorough discussion. Please see the attachment.
Cordially,
The authors.
Author Response File: Author Response.pdf
Reviewer 4 Report
The paper compares the performance of the network under different routing protocols. The contribution of the authors is not clear, in the section where the authors describe the enhanced version of OLSR protocol, they mention that they keep willingness on the maximum, in addition, they describe several hold times used in the protocol. The proposal to keep willingness on the maximum is not scientific contribution. The paper just investigates the performance of OLSR protocol under different protocol parameters, varying them in range allowed by the specification.
Moreover, there is no methodology in the paper, the simulation parameters are not clear and not sufficient, so it is not possible to reproduce the experiment.
For more comments see the attached file.
Comments for author File: Comments.pdf
Author Response
Dear Reviewers and Editors,
First of all, we (the authors) want to express our gratitude to the reviewers and editors for the thoughtful and helpful comments on an earlier draft of our manuscript, which is entitled “Comparative Evaluation of Enhanced Optimized Link State Routing with Topology-based Routing Protocols for UANET”. Your letter indicated that our manuscript (ID: drones-1510388) must be substantially revised to be considered for publication. We have made nearly all of the requested edits and feel as though this revised manuscript is vastly improved and ready for publication. What follows is a detailed list of the concerns and how we addressed them.
The response letter is attached in a file with a thorough discussion. Please see the attachment.
Cordially,
The authors.
Author Response File: Author Response.pdf
Round 2
Reviewer 1 Report
- A better description of the proposed E-OLSR protocol has been added, incl. the characteristics in OLSR routing table as well as some extra explanations have been given.
- For the simulation modeling in OPNET the authors have given a few comments, however there is still a lack of some facts/figures exposing the essence of such modeling. For more visibility fragments and/or configuration parameters as well as a screenshot with logs/results of such modeling would be presented.
- As for the UAV scenario the authors have given some additional details covering the system model and particular physical characteristics. However it'd be nice also to see the exact model and manufacturer of the used/modeled UAV.
- As for the experimental results my initial comment would remain valid still. It's nice the authors measure several target parameters, but I'd expect to see also any more valuable conclusions about these values, including ones on the applicability for the real cases reasoning these numerical values, anything on the effectiveness, etc.
- Concerning UANET vs. MANET some responses clarifying these aspects have been added to the article.
- For the Data Drop Rate a due calculation formula has been added.
- Needed explanations on the abbreviations have been added to the text.
- Figure 2 has been removed.
- Figure 4 has been improved.
- A few misprints has been corrected.
Author Response
Please see the attachment
Author Response File: Author Response.pdf
Reviewer 2 Report
Author Response
Please see the attachment
Author Response File: Author Response.pdf
Reviewer 3 Report
The authors correctly addressed all my comments.
Author Response
Please see the attachment
Author Response File: Author Response.pdf
Reviewer 4 Report
I am still insisting that the paper just investigates the performance of OLSR protocol under different protocol parameters, varying them in range allowed by the specification. This is not a new protocol (E-OLSR) that authors put as part of their contribution. Based on this observation I would recommend to amend the abstract, contribution, conclusion and title of the paper to show that OLSR performance can be improved for UANET by optimising the parameters.
The authors added the description of how their proposed protocol differs from OLSR. I have read the references that they have presented in order to prove that tuning of parameters can be considered as a new protocol. The papers did not convince me: the authors of one of the papers [34] used additional energy-based parameter to calculate willingness, unlike the authors of this manuscript that just decided to put willingness of nodes to maximum without any justification or comparative performance evaluation. Another paper [35] is telling about optimization of OLSR for VANET by tuning the parameters of OLSR from RFC 3626 but they did not call this optimization framework as a new protocol.
The paper [14] T. K. Priyambodo et al, ”Performance Optimization of MANET Networks through Routing Protocol Analysis”, Computers, 2021 is a very good example of the paper that uses tuning of the parameters and the authors call it performance optimization.
I agree that some of the OLSR parameters after optimization can outperform the standard ones defined in the RFC 3626 but this is absolutely not a new protocol.
You said correct - In UAVs willingness is crucial, as many nodes do not want to send traffic due to preserving energy – and in the proposed method you set up it to the highest value without considering energy capacity of the nodes.
The contribution of authors should be changed:
- The optimization of configuration parameters of the OLSR protocol to make it suitable for UANET with renewable energy.
- proposed OLSR configurations outperform the standard one
Line 320-327: Read again this paragraph, misprints here. The UAVs altitude is considered to keep constant between 200m and 50m. How the constant can be between 200 and 50m? It is variable then
The UAV’s speeds are selected as 40 ms and 224 30 m s, respectively. Respectively to what?
Extended rate PHY (802.11g) is used as the underlying MAC layer with a data rate of 24 Mbps and 1 Mbps, respectively. Respectively to what? Two experiments? But you have 4 scenarios.
Table 4 with simulation scenarios – I would move this table to methodology section or to the beginning of 4.1 Simulation Setup where you are talking about nodes.
I liked your additional information that were put into the text: methodology, parameters description, Table 4 with scenarios, missing input parameters for simulation (transmission power and buffer size). One very important comment about not considering Energy constraints (assuming that nodes are not battery powered and have no limit of energy) was added as well. I would highlight it in the abstract, intro and conclusion as well.
Author Response
Please see the attachment
Author Response File: Author Response.pdf
