Group data freshness scheme for outsourced data in distributed systems

Abstract

Data freshness ensures accessing recent data that could help in achieving high business values and providing effective customer service. Group data freshness is a challenging aspect in a distributed outsourced environment, as stale data among different entities may mislead the business goal of the system. Generally, three-party data outsourcing model is found in practice: users, data owner, and�cloud service provider. The users require to register with the data owner to access data files directly from the cloud service provider. A scheme verifying the freshness of whole�outsourced data�of its readers is called the group data freshness auditing scheme (GDFAS). Existing GDFASs focus on a probabilistic guarantee and require high computational cost at the data owner. In this paper, an efficient group data freshness auditing scheme is proposed, where the data owner does auditing in a distributed system with the help of the system users. As the data owner is not directly involved in their user�s data access, it needs mechanisms such as auditing data through an additional third-party to ensure the data is fresh. However, the third-party data storage service provider may not be fully trusted by a data owner. In such context, auditing data with respect to its freshness property without involving additional third-party storage service is challenging, but would be more effective in terms of the system�s performance and efficacy. The proposed GDFAS provides real-time data freshness verification using�Merkle hash trees. In comparison to the existing scheme, it takes less computational cost at the data owner without involving any third party and less communication cost between the data owner and the service provider. The proposed GDFAS is implemented on the AWS cloud and the auditing cost at the data owner is experimentally evaluated. The proposed GDFAS is analyzed and compared with the relevant existing scheme and is found that the proposed GDFAS outperforms other schemes with respect to its security and efficiency.