Critique of A Novel Proof-of-Reputation Consensus for Storage Allocation in Edge Blockchain Systems

Posted 2022-08-05 timerring

Critique

Paper title: A Novel Proof-of-Reputation Consensus for Storage Allocation in Edge Blockchain Systems

Authors: Jiarui Zhang, Yaodong Huang, Fan Ye, Yuanyuan Yang

Conference/Journal: 2021 IEEE/ACM 29th International Symposium on Quality of Service (IWQOS)

Critique Authors：timerring

Date: 2022.7.15

1. Summary of the Paper

This article proposes a novel proof-of-reputation consensus for storage allocation in edge blockchain systems. The article mainly analyzes the three aspects of reputation mechanism, storage allocation algorithm and blockchain improvement. The goal of optimization is quality of service (QoS) in edge network storage allocation. .First, for the reputation mechanism: each node publicly broadcasts the personal reputation list to all other nodes for evaluation, and each node can calculate the global reputation of all nodes by aggregating personal reputations. .Secondly, for the challenges in storage allocation in edge networks, the author adopts Fairness Cost (FDC) and Range-Distance Cost (RDC) to optimize the objective function, and proposes a storage allocation algorithm to ensure the reliability of storage allocation. the integrity and integrity of the requested data.. Finally, the article designs a PoR blockchain through the reputation mechanism and storage allocation algorithm. And in the worst case, the Decentralization Test, Fairness Test, Efficiency Test and Successful Access Rate Test were carried out. The results show that the proposed system has obvious improvement compared to the baseline.

2. comments on the paper

Read the full article. The advantages of the article are as follows:

1) The PoW mechanism commonly used in blockchain has the disadvantage of excessive consumption of computing resources, and PoS has the disadvantage of tending to be centralized. Therefore, the improved storage mechanism based on PoR in this paper avoids the above two shortcomings and ensures the blockchain’s performance. Decentralization and reliability.

2) The article proposes a better reputation mechanism and uses the EigenTrust algorithm, which can make it easier for each node to maintain a consensus list of reputations.

3) In the allocation of storage resources, the reputation mechanism helps nodes maintain the reliability of others, thereby helping to adjust the FDC and RDC in resource allocation to improve reliability and fairness.

At the same time, there are inevitably some flaws in the article：

1) To avoid malicious nodes that can appear malicious at the beginning and then honestly exhibit fluctuations in their global reputation between two different blocks, the article adds a constraint that block generators must have in the top 50% Global reputation does not strictly state the source of this 50% threshold and whether it is optimal.

2) In terms of algorithm evaluation, the baseline selected in this article is the PoS model in the reference Resource allocation and consensus on edge blockchain in pervasive edge computing environments, both of which have conducted Efficiency Test and Successful Access Rate Test, but only in Decentralization Test and Fairness Test. Self-control is set, and it is difficult to be convincing without comparison with other models.

3) In the Security Analysis section, the article describes the security of the model in narrative language, without experiments or data explanations, and it is not very convincing.

3. Suggested Improvements

1) The writing and review of the article are not rigorous. In section VIII, there is a typo: We have constructed a PoS blockchain based on our reputation mechanism. Combining with the full text, we can see that this is a PoR blockchain instead of PoS, which should be corrected.

2) The authors should account for the source of the 50% threshold and verify that it is optimal.

3) Compare with other PoR models in Decentralization Test and Fairness Test to enhance persuasiveness.

4) Set up an experiment during Security Analysis, and use data to illustrate its security level.