Reliably Controlling Massive Traffic between a Sensor Network End Internet of Things Device Environment and a Hub Using Transmission Control Protocol Mechanisms
Round 1
Reviewer 1 Report
Comments and Suggestions for AuthorsThe biggest weakness of the paper is that its context is very poorly described. In vain I searched for a diagram to show me what the paper was about. Where are the end points of the TCP connection? I assume it's not at at the sensor level but where? A Raspberry Pi is mentioned. A hub is also mentioned. Is it that the traffic from the sensors are aggregated at the Raspberry Pi and transported via TCP to the hub? Where is the hub? What assumptions are made about cross traffic on the TCP connection? About capacity? About reliability?
The second big issue with the paper is that it covers an area that has been the topic of decades of research - TCP parameter tuning. TCP has been examined, modified and experimented with in any number of scenarios. What is it about this paper that is novel? What does it contribute that thirty plus years of experimental and theoretical research have not yet uncovered. Modelling it as a queuing system is not novel.
The third issue is that they are using this approach to determine the appropriate window size for TCP. My understanding of TCP is that it uses the window size to control the average rate of traffic entering the network. The window size is addititvely increased until loss occurs upon which it is multiplicatively decreased. The nature of TCP is that it continually probes to increase the window size so it can adapt to network conditions. It transmits at full rate until the window is filled. It then waits for an acknowledgment. If received then the window size is increased and the process continues until there is loss. There may be intermittent cross traffic or other flows to the same destination. TCP window algorithm adapts to them. Why is it necessary to calculate the window size? What is the point of the research?
Finally, the paper appears to be an extension of an MDPI Electronics paper (reference 7). It would be good if the Introduction made this clearer. At the moment the paper says "The authors propose the integration of a sensor network end IoT device (SNEIoTD)". I was left wondering "what authors"? It wasn't until I looked at the reference that I realised it was the authors of this paper.
There are lesser issues as well. Section 1.3 is quite vague. I think the subsection title is odd "Research Basics". Is it a tutorial on doing research? Is it an overview of what the paper is exploring? Regardless, the title and the subsection needs thorough revision.
However, as noted, my main criticism is that before the paper can be accepted a lot more work needs to be done on explaining its context, its novelty, and its significance.
Comments on the Quality of English LanguageThe English needs work. Abbreviations such as "let's" are not appropriate in a journal article. Line spacing seems to change at random intervals. Section 2.1 is written as a long series of dot (dash) points. Either write it as simple prose or put the items in the list in a table.
Author Response
Dear Colleague,
please read the attached file.
Sincerely, Authors.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
Comments and Suggestions for AuthorsThe paper can be interesting for the future developments of IoT networks.
I address some comments to the authors.
The language should be carefully reviewed. Some statements are not clear and difficult to read. I suggest the reduction in the use of phrase edging, as some statements can be made more straightforward.
Section 1 has two different text spacings.
" As we mentioned 58 above, sensor networks are designed to collect critical information" What is the definition of critical information? Should this method be aimed to mission critical sensor networks? the authors should make this consideration with some examples.
Line 72 has a typo (two dots).
Section 1.2 - State of the art - can be expanded. Reporting the general structure of an IoT WSN would make this introduction section more complete and readable for the user, maybe introducing a block scheme of this general architecture.
The principle of SNEIoTD can be illustrated with a scheme too.
To the reviewer, the importance of TCP in IoT chains is clear, but it should be made clearer for non-expert readers. To this extent, some real scenario examples of WSN applications that include TCP in the data transmission chain can be introduced, some examples are reported here [A] [B]
[A] https://ieeexplore.ieee.org/document/9816747
[B] https://www.mdpi.com/2079-9292/12/2/457
Line 136, the TCP variants should be referenced.
Section 2
A table would be useful for associating the meaning to each symbol.
In section 3, figure 1, the legend is hiding a plot point. Line 428 can be moved before fig 1 and 2 to better explain those.
The line spacings need formatting also in this section.
Section 4 reports some repetitions of the abstract. The authors should shorten this section and be more concise about the obtained advancements.
"As a promising direction for further research" the bold formatting should be removed.
The authors mention a calculator device "Raspberry Pi" in the abstract as the base for a SNEIoTD, how does this influence the mathematical model? If it does not, this information is not relevant and can be removed.
The paper can benefit from a language review. Some phrases are very long and not clear.
Author Response
Dear Colleague,
please read the attached file.
Sincerely, Authors.
Author Response File: Author Response.pdf
Round 2
Reviewer 1 Report
Comments and Suggestions for AuthorsThe authors have explained the context of the work much more clearly. However the contribution of the work is still unclear. They still need to explain how this work is going to benefit anyone designing or developing IoT systems.
Comments on the Quality of English LanguageThe English needs work. “Dear reader” is not appropriate.
Author Response
Dear Colleague,
please read the attached file.
Sincerely, Authors.
Author Response File: Author Response.pdf
Reviewer 2 Report
Comments and Suggestions for AuthorsThe reviewer kindly thanks the authors for answering the previous comments.
Here are some minor comments to this version of the paper.
Line 58: "Dear reader. the endpoints in the studied system are a SNEIoTD, which aggregates information from sensors that generate massive traffic, and, as it accumulates, transmits it to the hub"
There is a typo in the first part.
SNEIoTD is singular (sensor network end IoT device) and the author refer to it it with plural in this case. The language should be checked again.
I suggest that the "Dear reader" can be omitted.
Fig.1 can be made clearer adding what type of connection (for example, TCP and where) is used among blocks.
In reference to the previous review comment: "5. Line 136, the TCP variants should be referenced."
The reviewer still thinks that these variants should be properly referenced in the text. Readers from other expertises should be able to follow the paper without deep knowledge of these TCP variants, or, at least, be able to document about those.
The suggested works for the introduction were aimed to the expansion in the general field of WSNs applications and real scenarios, as those rely on TCP for the intermediate and final transmission chain. Moreover, the suggested papers don't deal with the same topic as the text now suggests. It would be better to introduce this part as a general WSN scenario, possibly mentioning how TCP is used in the transmission chain.
The described logic of control (section 2.1) would be very well suited in an additional flow chart figure.
Comments on the Quality of English Language
Very small corrections to the language can be adopted, as suggested in the review comments.
Author Response
Dear Colleague,
please read the attached file.
Sincerely, Authors.
Author Response File: Author Response.pdf
