Session Chair: Sharding Blockchain
• Dr. Liang Cheng, Lehigh University, USA
Sherman S.M. Chow
The Chinese University of Hong Kong, Hongkong, China
Introduction: Ever since the success of Bitcoin, blockchain technology has enabled countless decentralized applications. However, one major issue of blockchain is scalability. Taking Bitcoin as an example, its maximum transaction throughput is estimated to be less than 10 transactions per second, which is restricted by the constant block size and the constant average confirmation time per block of the Bitcoin blockchain. Some other cryptocurrencies that use similar blockchain technology, such as Ethereum and Bitcoin Cash, increase their throughput by enlarging these two constants. Yet, none of them achieves similar throughput of mainstream payment networks like Visa (over 1500 transaction/s) or Paypal (~200transactions/s). A more serious problem is that we cannot increase these two parameters arbitrarily. Doing so may damage the underlying decentralized consensus protocol, as well as increase the cost of running full nodes. The performance bottleneck hinders blockchain’s applicability, especially in resource-constrained platforms such as IoT.
Sharding, originally a classical database design principle, is now being considered as a promising way to solve the scalability issue of blockchain protocols. It works by distributing the overhead of validation and/or storage of transactions to multiple groups of nodes that work in parallel. Therefore, the more groups in the network, the higher the throughput. Yet, the current blockchain sharding protocols, namely, Elastico (CCS ’16), OmniLedger (S&P ’18), and RapidChain (IACR ePrint ’18) have respective weaknesses in practicability, performance, or security. Elastico is known to have high failure probability and high latency. OmniLedger is known to have prohibitive communication cost in Byzantine setting, and the underlying consensus protocol, ByzCoin (USENIX Security ’16), has several security issues (DISC ’17, OPODIS ’17). RapidChain, the most recent proposal, uses a synchronous consensus algorithm. The assumption of a synchronous channel is strong for a peertopeer blockchain system. Finally, sharding intrinsically introduces many cross-shard transactions that require costly two-phase commit style protocol to handle, which is not needed in the non-sharded blockchain.
Here, we call for a full sharding blockchain protocol that supports both UTXO model and account-based model, admits rigorous security analysis, and most importantly have comparable throughput with state-of-the-art commercial payment systems.
Biography: Sherman S.M. Chow joined the Department of Information Engineering at the Chinese University of Hong Kong as an assistant professor in November 2012 and received the Early Career Award 2013/14 from the Hong Kong Research Grants Council. He was a research fellow at Department of Combinatorics and Optimization, University of Waterloo, a position he commenced after receiving his Ph.D. degree from the Courant Institute of Mathematical Sciences, New York University. During his study, he interned at NTT Research and Development (Tokyo), Microsoft Research (Redmond), and Fuji Xerox Palo Alto Laboratory.
His main interests are in Cryptography, Security, and Privacy, with publications in CCS, EuroCrypt, ITCS, NDSS, and USENIX Security. He served on the program committee of AsiaCrypt in a consecutive of six years, and of 150+ other conferences including CCS, ESORICS, ICDCS, Infocom, PKC, PETS, and TheWeb. He is a program co-chair of Blockchain and Sharing Economy Applications, co-located with IEEE ICDM 2018, and CANS, AsiaCCS-SCC, ISC, and ProvSec before. He also serves on the editorial boards of IEEE Transactions on Information Forensics and Security (TIFS), IET Information Security, etc.
In the areas of blockchain and cryptocurrencies, he chaired the session of Payments and Security in CCS 2014, served on the program committee of Financial Cryptography and Data Security in 2013 and 2015, and has been a program committee member of ACM Workshop on Blockchain, Cryptocurrencies and Contracts since its inauguration.
Lancaster University, UK
Utilizing IoT Devices to Authenticate Out-of-Band Data in the Blockchain Context
Abstract: The blockchain technology pioneered a new approach to build a global distributed public ledger that is persistent and immutable. It can be used to store and process (e.g. via smart contracts) information from the external world, which we call out-of-band data. An oracle is a third-party service that provides out-of-band data to the blockchain context. The integrity of the fed data largely relies on the trust of the corresponding oracles. For instance, blockchains are widely used to enhance traceability and transparency of the physical supply chain nowadays. Although the data on the blockchain is globally visible, anything that happens outside of the ledger remains in the dark. As a result, any fraud or malicious collude during the initial data supply process in the physical world may lead to catastrophic end-to-end traceability breaches.
In this talk, we introduce a new type of information oracle that is based on the consensus of a randomly sampled collection of IoT devices. Suppose each IoT device has an identity (public key) registered on the blockchain. During the data dismissing phase, a random subset of those IoT devices are selected based on the concept of cryptographic sortition proposed by Micali. The selected IoT devices will jointly participate the decentralized decision-marking process to propose and certify certain out-of-band data. In particular, the system enables liquid democracy to achieve better collaborative intelligence. Liquid democracy (also known as delegative democracy) is an hybrid of direct democracy and representative democracy, where the participates can either directly express their opinions or delegate their rights to representatives who vote on their behalf. It offers the benefits of both systems (whilst doing away with their drawbacks) by allowing the IoT devices to take advantage of those confident data sources.
Biography: Dr. Bingsheng Zhang is a lecturer (a.k.a. assistant professor) in the School of Computing and Communications at Lancaster University and a core member of the Security Lancaster, one of the first 8 Academic Centres of Excellence in Cyber Security Research recognized by EPSRC and GCHQ, UK. He is the leader of the security group at Lancaster University and the program director of Msc. Cyber Security. He has several EPSRC and PETRAS grants on e-voting security and blockchain security. He also received over GBP 0.6M donation from various world’s leading blockchain R&D companies, such as the Blockchain Institute and IOHK. Dr. Zhang specializes in cryptography and cyber security. He has published over 40 papers in the world’s leading top-tier security/cryptography conferences and journals, such as EUROCRYPT, ASIACRYPT, PKC, SCN, ACM CCS, ICDCS, PODC, Financial Cryptography, INFOCOM, IEEE S&P, IEEE TMC, and IEEE TIFS. In the recent years, his research efforts mainly focus on end-2-end verifiable e-voting and blockchain security.