A Comparative Study of Consensus Mechanisms in Blockchain for IoT Networks

Round 1

Reviewer 1 Report

The authors propose a comparative study of blockchain consensus for IoT networks. The authors have organized the paper well. However, the authors must make additional improvements to improve the quality of the paper.

My main comments:

This manuscript is too general and lacks details such as taxonomies, comparisons, and diagrams that mainly exist in a high-quality survey paper. Hence, readers with a basic understanding of blockchain consensus can only gain a little insight after reading this manuscript.

Further details are necessary to be added. I suggest the authors search other "blockchain consensus survey" papers for inspiration.

My other comments:

The abstract should also mention about findings/output from the research.

Move the paper's contribution from Section 2 to Section 1.

I think that Sections 3 and 4 are unnecessary since they explain general things about the blockchain. Since this paper discuss an advanced topic (consensus of blockchain), the readers of this manuscript are expected to know about blockchain already.

The separation between "Open blockchain" and "Case Study consensus" is confusing. Since the given "case study consensus" depends on the existing consensus, such as PoW and PoS, they can be considered as "variants" or "sub-variants" of the mentioned consensus. Hence, I believe they should not be separated as case study consensus.

Because this is a survey paper, I understand that the authors cannot explain the details of each consensus in this paper. Instead, interested readers should refer to the original paper to understand the consensus. However, I think it would be better if the authors made some effort to explain the general flow of each consensus. For example, the authors can make a figure for each consensus, such as in Figure 3 and Figure 4. Therefore, each type of consensus has its own figure, which shows step-by-step how the consensus works. The current explanation in the paragraph is too general.

Why are the comparison item in Table 1 and Table 2 different? I think they should be compared equally.

The authors mention L1 in Table 3 and L2 in Table 1, but there is no definition of L1 and L2 in the paper.

Table 3 still needs reference numbers in each row for quick access (even though the contents are similar to Tables 1 and 2). Furthermore, what does "Storage" means in this table? What do check, cross, and dash mean? The authors should clearly describe the table's content.

In section 7.4., how do the authors determine that the given consensus is suitable for IoT? For example, in Table 3, PoL is better at computing, security, decentralization, and throughput. From the authors' scoring, seems to me like this consensus is good at solving blockchain trilemma, but why is this consensus not suitable for IoT?

Similarly, why are use case consensus tailored explicitly for IoT use cases (PoSCS, CGPoW, PoEWAL, and Microchain) unsuitable for IoT? The authors should clearly describe their reasonings and justifications.

In 7.4.1., the authors recommend PoS if decentralization is vital in their system. However, PoS tends to be centralized (having less number of block generator nodes compared to, e.g., PoW).

Add a conclusion section, where the authors summarize the paper's goal, methodology, and research results/output.

Does blockchain with capital B necessary? If yes, then must make all blockchain to Blockchain for consistency.

 

Author Response

Thank you for taking the time to thoroughly review our paper.

Please see the attachment for our response.

Author Response File: Author Response.pdf

Reviewer 2 Report

The paper describes and analyses in-depth consensus algorithms for blockchain adoptable by IoT end nodes. Apart from consensus mechanisms thought for traditional blockchain, the paper analyses ad-hoc consensus algorithms designed for IoT. 

My primary concern is the following. The paper assumes that the blockchain nodes are the IoT sensors. However,  this is not recommended because IoT sensors are very constrained devices. Indeed, they are battery-powered, have scarce network bandwidth, cannot always be online, can execute easy computational tasks, and so on. In [1], the authors describe the most adopted architecture solutions to integrate the blockchain with an IoT ecosystem. It would be interesting to see how different consensus algorithms combine with the various IoT-BC integration architectures. 

Other comments:

* The authors introduce the criteria in section 2. However, these are not explicitly considered in the analysis section, in table 2 and 3. To clarify more, I suggest doing this.

* Are there other papers analysing consensus on IoT devices? If yes, how this paper differs from them?

* The paper is missing the conclusions to summarize the paper's analysis. What are we taking away from the paper? Which are the open problems? I suggest including this in section 8.

In my opinion, the paper should be considered for publication after addressing these comments.

[1] Pennino, D.; Pizzonia, M.; Vitaletti, A.; Zecchini, M. Blockchain as IoT Economy Enabler: A Review of Architectural Aspects. J. Sens. Actuator Netw. 2022, 11, 20. https://doi.org/10.3390/jsan11020020 

Author Response

Please see the attachment

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

The authors have addressed all of my concerns in the updated manuscript. Still, there are a few formatting issues that need to be fixed:

 

- In Line 50-51, "we 'play' close attention.....?"

- The contribution in Section 1.2. should be made in numbering or bulleting lists.

- In Line 108, 119, 138, and so on..., the beginning of the paragraph should have an indentation.

 

Author Response

Please see the attachment.

Author Response File: Author Response.pdf