An IOTA-Based Service Discovery Framework for Fog Computing

Round 1

Reviewer 1 Report

The paper addresses the implementation of a decentralized service discovery based on a IOTA ledger for Fog Computing

The research is clearly defined. Also, the cost/performance analysis in chapter 4 proves the solution can be used for common use-cases.

Maybe the authors can address/highlight how they will provide the authentication of valid service/content providers that will register they service in the ledger. Someone can register rogue nodes close to the victim in order to force the later to connect to it.

I have no other major observations.

Author Response

Responses to reviewers' comments

Reviewer 1:

1. Maybe the authors can address/highlight how they will provide the authentication of valid service/content providers that will register they service in the ledger. Someone can register rogue nodes close to the victim in order to force the later to connect to it.

Response: The IOTA Foundation provides digital certificates to all participating parties, and those parties initially trust the DID (Decentralized Identifier) of the IOTA Foundation. Eventually, these participants, such as service requestors or providers, can start to create digital certificates for their actors, growing their network of trust towards an increasingly decentralized trust model. The methods for creating, reading, updating, and deactivating the DIDs and the associated DID documents are described in the respective specification (DID method)[33]. We have added the above description in our future work on page 18 to explain how we will provide the authentication of valid service providers.

Reviewer 2 Report

This article studies an important and recent topic. The authors organized this research well. They provided good Introduction and methodology Sections as well as described appropriately evaluating the performance of their framework, but it requires some improvements. However, the authors must accurately address the below comments.

• Abstract: The authors clarified the problem and the purpose of this study. The solution methods are also well illustrated. However, the findings of this study are not explained briefly at the end of the abstract. The abstract should be comprehensive to be clear to readers.
• Keywords: We suggest that the authors should remove keywords such as “fog computing”, "service discovery" and “IOTA” because these keywords are already found in the review article title. It is better that they replace them with other keywords to increase the reach of the article.
• Some acronyms should be defined before using such as AWS, CDN, DHT, ... etc.
• We think that the authors should add a subsection (in 4. Performance Evaluation) that includes a comparison of the proposed framework with the existing one, which demonstrates the superiority of the proposed framework over the previous schemes (it is preferable to compare the results obtained with the results of the related research). It is also suggested to clarify limitations in the proposed framework.
• The authors pointed out that their framework provided higher security than DHT to defend against cyber-attacks such as Sybil and Eclipse. However, they did not explain how their framework is capable of fending off these attacks and furthermore they did not provide an analysis of the security issues.
• Figures and Tables: All figures and tables are drawn with high resolution. However, Figures 3, 4, 12, 14, 16 and 17 are shown before their use in-text.
• Paraphrasing: Authors must paraphrase all sentences and paragraphs taken from their previous research, for example, paragraphs "In terms of implementation, IOTA reference implementation (IRI) has been ... and stored in the database and then is for-219 warded to other neighbors. (page5-lines213-220)" and "When the client wants to initiate a transaction, the process is described as ... IRI for verification, storage, and broadcasting to complete the transaction initiation. (page5-lines221-229)" must paraphrase (check the entire article). Although this research belongs to the same authors, the authors must avoid taking whole sentences from the original papers. Also, their previous paper should include in the list of references. The authors are obligated to paraphrase during the revision of the entire article.
• Proofreading: This article needs minor proofreading such as remove “\” (page2-line95), remove “(genesis)” (page4-line197), remove "(Masked Authenticated Message)" (page5-line235) because it is a duplicate. Some paragraphs are long, such as the first paragraph page 3 and the second paragraph page 4.
• List of References: The references should follow Electronics-MDPI style. It requires extensive improvement. Researchers' names in the reference list incorrectly wrote, such as [1], [2], [3],[4], … etc., some search names in the reference list begin an uppercase letter for each word (such as [16], [17], [19], ... etc.) and others use only an uppercase letter in the first word (such as [8], [9], [10], … etc.), authors should standardize style. The double quote should remove search names such as [1], [2], [3], … etc., some journal names are not italic such as [1], [9], …etc., some of the references do not contain enough information such as [6], [7], … etc. Author must accurately check list of references to remove all problems.

Author Response

Responses to reviewer’s comments

Reviewer 2:

1. However, the findings of this study are not explained briefly at the end of the abstract. The abstract should be comprehensive to be clear to readers.

Response: Thanks for the reviewer comment. We have added the finds from our experimental results into the abstract on page 1.

2. Keywords: We suggest that the authors should remove keywords such as "fog computing", "service discovery" and "IOTA" because these keywords are already found in the review article title. It is better that they replace them with other keywords to increase the reach of the article.

Response: As shown on page 1, we have changed the keywords as distributed ledger technology, masked authenticated message (MAM), time-difference registration addressing, tree-based MAM indexing, and node selection to highlight the developed and used for the implementation of the proposed framework.

3. Some acronyms should be defined before using such as AWS, CDN, DHT, ... etc.

Response: We have defined the acronyms before we use them in this paper.

4. We think that the authors should add a subsection (in 4. Performance Evaluation) that includes a comparison of the proposed framework with the existing one, which demonstrates the superiority of the proposed framework over the previous schemes (it is preferable to compare the results obtained with the results of the related research). It is also suggested to clarify limitations in the proposed framework.

Response: Thanks for the reviewer’s comment. However, we can not get the software of related work and estimate the performance of the previous schemes by executing actual experiments. We only can read the experimental results from the reference papers and compare the results with the performance of our proposed framework. For example, the experimental result of Blockchain-ased UDDI Data Replication and Sharing [19] shows that the service provider saves the detailed information of itself, and the service it provides to the blockchain is 17 seconds. The time that the service consumer gets the service-related information from the blockchain is 1 second. The time that the UDDI registry implements synchronous replication is 1 second. By contrast, our experimental results show that a fog node in the proposed framework needs to spend only 2 seconds to finish the creation and storage of a MAM transaction with accelerator support and spend only 450 ms to propagate the MAM transaction other nodes with no pay. The cost of query the information of known nodes is less than 200ms. This comparison shows that the proposed framework based on IOTA is practical for supporting service discovery.

5. The authors pointed out that their framework provided higher security than DHT to defend against cyber-attacks such as Sybil and Eclipse. However, they did not explain how their framework is capable of fending off these attacks and furthermore they did not provide an analysis of the security issues.

Response: We have added two paragraphs to explain how IOTA can provide adequate protection against Sybil and Eclipse attacks from page 4 to page 5.

6. Figures and Tables: All figures and tables are drawn with high resolution. However, Figures 3, 4, 12, 14, 16 and 17 are shown before their use in-text.

Response: We have improved these figures according to the reviewer's suggestion.

7. Paraphrasing: Authors must paraphrase all sentences and paragraphs taken from their previous research, for example, paragraphs "In terms of implementation, IOTA reference implementation (IRI) has been ... and stored in the database and then is for-219 warded to other neighbors. (page5-lines213-220)" and "When the client wants to initiate a transaction, the process is described as ... IRI for verification, storage, and broadcasting to complete the transaction initiation. (page5-lines221-229)" must paraphrase (check the entire article). Although this research belongs to the same authors, the authors must avoid taking whole sentences from the original papers. Also, their previous paper should include in the list of references. The authors are obligated to paraphrase during the revision of the entire article.

Response: We have rewritten the paragraphs shown from page 5 to page 6 and have included our previous paper in the references [31] and [32].

8. Proofreading: This article needs minor proofreading such as remove "\" (page2-line95), remove "(genesis)" (page4-line197), remove "(Masked Authenticated Message)" (page5-line235) because it is a duplicate. Some paragraphs are long, such as the first paragraph page 3 and the second paragraph page 4.

Response: We have removed these words according to the reviewer's suggestion.

9. List of References: The references should follow Electronics-MDPI style. It requires extensive improvement. Researchers' names in the reference list incorrectly wrote, such as [1], [2], [3],[4], … etc., some search names in the reference list begin an uppercase letter for each word (such as [16], [17], [19], ... etc.) and others use only an uppercase letter in the first word (such as [8], [9], [10], … etc.), authors should standardize style. The double quote should remove search names such as [1], [2], [3], … etc., some journal names are not italic such as [1], [9], …etc., some of the references do not contain enough information such as [6], [7], … etc. Author must accurately check list of references to remove all problems.

Response: We have corrected the reference list by following the Electronics-MDPI style as shown from page 19 to page 20.

Author Response File: Author Response.pdf

Round 2

Reviewer 2 Report

Thanks, the authors have addressed my concerns.