北京科学与工程计算研究院学术报告之八十四

报告人/Speaker: 李泉林 (燕山大学)


报告题目/Title: Mathematical Models and Theoretical Challenges in Blockchain: From Reward Processes to Stochastic Processes


时间/Date & Time: 2018年11月13日,9:30—10:30


地点/Location: 理科楼M843


报告摘要/Abstract:
This talk focuses on mathematical modeling and analysis of blockchain, which have been developed as important basic theory of blockchain technologies in the past ten years. To this end, our analysis will provide some details from three aspects: Reward and incentive methods, consensus mechanism, and economic theory. We address five important issues related to mathematical models of blockchain:
(1) Physical topology, organization and structure for key mathematical issues of blockchain.
(2) Mining management and consensus mechanism. Analyzing the miner pool (or alliance), and crucially providing stochastic models to discuss double-spending and chain forks.
(3) From transaction fees, miner reward, difficulty of solving PoW puzzles, mining strategies, and consensus mechanism, establishing stochastic point processes, including inhomogeneous Poisson processes, and periodical Markov arrival processes.
(4) From interactions between transaction and block-generation arrivals, setting up matching and parallel-competing queues and stochastic networks, and more generally, Markov processes (including Brownian motion) under some specific consensus mechanism. Based on this, we can achieve performance evaluation and optimization of blockchain systems. In addition, from attack strategies and chain forks, developing Markov decision processes, and game models including non-cooperative game, cooperative game, evolutionary game and stochastic game.
(5) From economic theory, together with complex networks, this talk further discusses basic role and driving force played by blockchain technologies, which greatly motivates future economy of our global world. Here, we emphasize several key aspects: Cost decomposition of blockchain applications, technological innovation across different fields, and market driving force of relevant products and their deep exploitation.


报告人简介/About the speaker:

李泉林,博士,教授、博士生导师。1998年在中国科学院应用数学研究所获得博士学位;1999年7月到2003年12月为中国科学院自动化研究所模式识别国家重点实验室副研究员;2003年12月到2009年10月为清华大学工业工程系副教授;2009年10月到现在为燕山大学经济管理学院教授、博士生导师。李泉林已经在国际学术刊物上发表了60余篇SCI学术论文,其中SCI索引500余次、他人SCI索引400余次,他20余次担任排队论、随机模型与应用概率等领域重要国际学术会议的学术委员会主席与委员。他已经获得了2004年教育部新世纪优秀人才、2005年教育部科学技术奖(自然科学)一等奖、2007年北京市科学技术(自然科学)二等奖、2008年北京市精品课、2013年河北省高等学校科技领军人才、2014年河北省科学技术(自然科学)二等奖、2015年国际INFORMS优秀论文奖。他已经主持并负责了20余项国家973计划、国家863计划、国家自然科学基金、国家自然科学重点基金和国内外大型企业的合作项目。