面向去中心化存储的数据流行度去重模型

数据流行度去重方案中存在检测机构不诚实、数据存储不可靠等问题，提出一种面向去中心化存储的数据流行度去重模型。针对检测机构不诚实，模型结合区块链的不可篡改性与智能合约的不可抵赖性，将智能合约作为检测机构执行数据的重复性检测和流行度检测，保障了检测结果的真实性。针对数据存储不可靠问题，提出一种文件链存储结构，该结构满足数据流行度去重的要求，并通过添加辅助信息的方式，建立分布在不同存储节点中实现物理/逻辑上传的分片之间的逻辑关系，为流行度数据去中心化网络存储提供基础；同时，在数据块信息中添加备份标识，借助备份标识将存储网络划分为两个虚拟存储空间，分别实现数据和备份数据的检测与存储，满足了用户备份需求。安全性分析和性能分析表明，该方案具有可行性，保障了检测结果的真实性，并提高了数据存储的可靠性。

Data popularity deduplication model for decentralized storage

There are issues such as dishonest checking organizations and unreliable data storage in a deduplication scheme based on data popularity. This paper proposed a data popularity deduplication model for decentralized storage. To address the dishonesty of the checking agency, the model executed duplication checking and popularity checking of data by using smart contracts as the checking organization, which combined the non-tamper-ability of blockchain with the non-repudiation of smart contracts. It guaranteed the authenticity of the test results. To address the problem of unreliable data storage, it proposed a file chain storage structure. This structure satisfied the requirement of data popularity deduplication. Meanwhile, this structure established the logical relationship between the fragments which were distributed in different storage nodes to achieve physical/logical uploads by adding auxiliary information. The file chain storage structure provided the basis for popularity data in decentralized network storage. Besides, it added backup identifiers to the data block information, and divided the storage network into two virtual storage spaces with the help of the backup identifiers. Also, two virtual storage spaces were oriented for data detection and storage. It satisfied the user''s backup needs. Security analysis and performance analysis show that the scheme is feasible, guarantees the authenticity of the test results, and improves the reliability of data storage.