Theoretical Discussions on Blockchain-based Systems: Design, Extension, and Application

Abstract

Blockchain techniques have received an outburst of interest both in academia and industry, with numerous benefits such as decentralization, persistency, anonymity and auditability. There has been widespread adoption of blockchain in various fields ranging from cryptocurrency, financial services, risk management, internet of things (IoT) to public and social services. As a distributed ledger, blockchain records data in the form of linked blocks secured by cryptography. Instead of being controlled by a central third party like a bank, this ledger is maintained by a P2P network. To work properly, it is essential that the nodes in the network have the same local view of the ledger. A consensus mechanism is the core of a blockchain system since it regulates how to achieve the same states among the nodes in a distributed fashion. Specifically, it grants a qualified miner the right of appending the ledger in a period of time. Despite its potential as a disruptive technology, blockchain is currently facing multiple challenges and problems, which incurs some issues on the performance, security, and fairness of blockchain-based systems. There are many underlying reasons for those challenges and problems, such as an inefficient networking protocol, a time-consuming consensus algorithm, a strategic mining node and so on. In this dissertation, we summarize our theoretical studies on the above-mentioned topic from three different aspects. Details are as follows: Design: A major weakness of the current blockchain systems is the low transaction throughput, which is resulted from the small block size, the long block generation time, and the delayed block propagation. To shorten the block propagation delay, we design a neighbor selection algorithm, improving the communication speed in the P2P network of the blockchain. This design is an attempt on topology optimization and reduces the frequency of blockchain forking. Forking is a situation where the blockchain temporarily diverges into two or more branches, which wastes mining power and causes security issues. Extension: Although a consensus mechanism aims to regulate mining nodes’ behavior, it also can be manipulate by strategic miners within legal limits due to their profit-driven nature. In the dissertation, we analyze strategies that miners apply for utility maximization under different consensus mechanisms from the perspective of game theory. And then we propose distributed algorithms to optimize resource allocation among miners. Application: We also propose a blockchain-based market where we believe the application of blockchain will enhance the corresponding applications. It is a blockchain-powered data market that allows multi-user cooperative search. We also integrate smart contract into these applications. All discussions in this dissertation are theoretical explorations. The theoretical results may not be realistically practical since we hold strong assumptions. We hope that, we could fill the gap between theory and practice in the future.