UCLA

A blockchain characterized by a distributed, decentralized, and public computer system rapidly grows and realizes the Internet of Value. The fundamental idea is to replicate its state database to every node computer with protocol-specific restrictions to keep consistency and integrity. Transferring controls from a central entity to each individual sounds compelling, but a protocol designer should carefully design an incentive scheme to encourage anonymous participants to do the desired actions.

In this dissertation, we analyze the reward policy on block creation in a proof-of-work blockchain. The reward policy consists of a block reward and transaction fees. A protocol designer intends to incentivize block miners to process transactions until the designed capacity. However, we observe that the hidden costs may distort the reward policy; block miners do not have incentives if processing and network delays are not negligible. We model a transaction fee market, the block mining process with the delays, and the decision-making of block miners under the reward policy. We express the expected mining revenue as a function of the number of transactions in a block. We then identify the operating region of a block reward to mitigate the effect of the hidden costs. We propose a dynamic block reward of a proportional subsidy for processing transactions as a revised incentive scheme. Thus, blockchain users experience reduced waiting time and less expensive transaction fees.

In addition, we further expand queueing theory, especially for a head-of-line bulk service priority queue with an arbitrary service time distribution. We provide analytical expressions (exact solutions) for critical numbers in a transaction fee market: average waiting time, minimum transaction fee, and average total transaction fee.