Towards Secure and Scalable Blockchain Technologies


引用 0|浏览7
ABSTRACTThe blockchain technology is rapidly gaining grounds as a key technology, especially in the financial and supply chain management sectors. This is largely due to the ability of the technology to (i) efficiently manage the sharing of digital resources between a large number of stakeholders and (ii) to efficiently manage disputes arising in the process. In spite of its many advantages, experience with existing blockchain proposals reveals that there are still many challenges that need to be overcome prior to any large scale industrial adoption, namely [Scalability] Existing permissionless blockchains (e.g., Bitcoin) are able to scale to a considerable number of nodes at the expense of attained throughput (e.g., Bitcoin can only achieve few transactions per second[2]). On the other hand, permission-based blockchains can achieve relatively higher throughput, but can only scale to few hundred nodes. However, one needs to cater for both performance and scalability to meet industrial standards. [Privacy of lightweight clients] Most open blockchain platforms support lightweight clients, targeted for devices like smartphones, that only download and verify a small part of the chain. Here, clients connect to a full node that has access to the complete blockchain and can assist the client in transaction confirmation. As the full node has to learn all transactions issued and received by the requesting client to verify their correctness, such action obviously violates user privacy[1]. In this talk, we plan to overview a number of security challenges pertaining to existing blockchains-effectively capturing almost 8 years of research in this area of work. Moreover, we plan to discuss the performance limitations of existing blockchain-based consensus algorithms and explore different concepts leveraging trusted execution environments (TEEs) to enhance the scalability and security of existing consensus algorithms[3]. Finally, we will discuss the privacy provisions of existing lightweight client implementations and explore the solution space to enhance user privacy by leveraging functionality from TEEs[4].
AI 理解论文