Secure Data Distribution Architecture in IoT Using MQTT
Round 1
Reviewer 1 Report
This paper focuses in the MQTT protocol, and the problem for clients to subscribe to content provided by a broker. It is not clear why the protocol per se solves this issue, and the need to address the issue by a third party. At the hearth of the work, few is discussed about the secure distribution of content and the IoT connection.
The following are some comments about this paper:
1) First of all, the abstract can be significantly improved:
a. Null details of proposed architecture are now provided. What does exactly the architecture solve? What are the contributions?
b. Some details of the experiments must be commented as well as the description of what metrics are considered to evaluate the overhead.
c. Authors said that their results are promising, but nothing is said about the results and metrics evaluated.
d. “broker” is repeatedly mentioned in the abstract, but more context is needed.
2) A reference is missing by the end of line 36.
3) I suggest embedding Section 3 into Section 2, and provide a Table summarizing the state od the art. This table with qualitative data about related works could support the statements in lines 140 – 146.
4) I consider and suggest to better change bullets by numbers (enumeration instead of itemization)
5) Section 4 is very short and contains information already presented in the introduction. The section should be unified and embedded into Section 1.
6) Although authors state that the proposed architecture is secure, security is vaguely defined and considered in the design. It seems that authors only consider authentication service, but that task is delegated to a third party, not deployed in the proposed architecture. Thus, security is poorly addressed in this work. In fact, the security property (content) is not evaluated during experimentation.
7) Experimentation lacks some details. Was experimentation done in simulation? à what tools were used? Was it done in real IoT devices? à devices details are missing. Was the protocol implemented in some programming language or a reference implementation was used?
8) The scenarios for evaluation described from line 305 must be contextualized in an application scenario, a real application. For example, in what applications data published is 1 – 10000 bytes?, In what applications messages are sent 1 per second?
Author Response
Dear Ms. Pichayanin Jirathunyakorn,
First of all, we would like to thank you for your efforts and cooperation. We would like also to thank the reviewers for their time in reading our manuscript and providing valuable comments to improve the quality of our manuscript. Here, we are writing this to inform you on how we have addressed the issues raised by the reviewers. These changes have been incorporated in the revised manuscript.
Author Response File: 
Author Response.doc
Reviewer 2 Report
The paper proposes a novel architecture for MQTT protocol. The main advantage of the proposed architecture is usage of multiple distributed brokers.
The topic of the paper is interesting and of practical importance. I have some remarks about the presented results.
1) The authors state that the test were performed on the same machine. I think that they should be performed on at least two machines to include influence of network delay
2) The authors themselves state that using a single publisher could be the reason that payload does not have a major impact on delay. The impact of using at least two publishers should be investigated.
I recommend that the authors revise the article by expanding the experimental section according to the points above.
Author Response
Dear Ms. Pichayanin Jirathunyakorn,
First of all, we would like to thank you for your efforts and cooperation. We would like also to thank the reviewers for their time in reading our manuscript and providing valuable comments to improve the quality of our manuscript. Here, we are writing this to inform you on how we have addressed the issues raised by the reviewers. These changes have been incorporated in the revised manuscript.
Author Response File: Author Response.doc
Round 2
Reviewer 1 Report
Authors have satisfactory addressed my previous concerns. As a reviewer, I require a revised manuscript with changes tracked and more discussions on the responses provided. Please provide this material to proceed with the validation of changes done over the original submission.
Author Response
We would like to thank the reviewer for the useful comments and suggestions. Please see our revised manuscript with changes tracked. Our responses do not just include the line numbers where the changes are made but also show the changes tracked for each comment. Thanks.
