Vassago: Efficient and Authenticated Provenance Query on Multiple Blockchains
Rui Han, Jiang Xiao, Xiaohai Dai, Shijie Zhang, Yi Sun, Baochun Li, Hai Jin
Abstract
Recent successful pilot studies on blockchains showed significant benefits of data provenance towards improved visibility and authenticity, as well as a reduction of administrative costs. Traditionally, all the parties reside in a single blockchain system, which leads to various single-chain provenance query schemes. However, for the sake of convenience, it has been a trend for different parties to deploy their own blockchain systems separately, which requires cross-chain provenance queries. Unfortunately, state-of-the-art blockchain query schemes suffered from inauthentic transactions and low efficiencies when expanding to adversarial cross-chain commercial deals. To be more specific, query results may be inconsistent and incomplete over multiple blockchains due to the lack of global knowledge on cross-chain dependencies. Moreover, these schemes perform cross-chain provenance queries in sequence, leading to high query latencies. In this paper, we present Vassago, the first multi-chain system that empowers efficient and authenticated provenance queries. Vassago incorporates three innovative designs: 1) it explores the dependencies of cross-chain transactions, 2) it validates the authenticity of cross-chain provenance by recording and querying the dependencies on a shared blockchain, and 3) it improves efficiency by parallelizing query processes. Our experimental results show that Vassago can shorten the query latency by 85.9% and reduce the storage consumption by up to 85.7%, with negligible overhead.