Prof. David Starobinski (Boston University)
Wednesday, 18.2.2015, 11:30
Advance reservation (AR) services form a pillar of many branches of the economy, including transportation, lodging, dining, and, more recently, cloud computing. In this work, we use game theory to analyze an AR system in which customers differ in their lead (arrival) times and the number of customers requesting service is random. Based on statistical information, the customers decide whether or not making an advance reservation of server resources for a fee. We prove that only two types of Nash equilibria are possible: either none of the customers makes AR or only customers with lead time greater than some threshold make AR. Our analysis further shows that the fee that maximizes the provider's profit may lead to other equilibria, one of which yielding zero profit. In order to prevent ending up with no profit, the provider can elect to advertise a lower fee yielding a guaranteed, but smaller profit. We refer to the ratio of the maximum potential profit to the maximum guaranteed profit as the price of conservatism. We prove that in the single server case the price of conservatism equals one, but can be arbitrarily high in a many-server system. Finally, we analyze the dynamics of AR games based on two learning models: action-learning and strategy-learning. We show that if the provider is risk-averse then action-learning yields higher profit, while if the provider is risk-taking then action-learning yields zero profit in contrast to strategy-learning.
David Starobinski is a Professor of Electrical and Computer Engineering at Boston University, with a joint appointment in the Division of Systems Engineering. He is also a Faculty Fellow in the US DoT Volpe National Transportation Systems Center. He received his Ph.D. in Electrical Engineering from the Technion-Israel Institute of Technology, in 1999. In 1999-2000, he was a visiting post-doctoral researcher in the EECS department at UC Berkeley. In 2007-2008, he was an invited Professor at EPFL (Switzerland). Dr. Starobinski received a CAREER award from the U.S. National Science Foundation (2002), an Early Career Principal Investigator (ECPI) award from the U.S. Department of Energy (2004), the best paper award at the WiOpt 2010 conference, and the 2010 BU ECE Faculty Teaching Award. His research interests are in the modeling, performance evaluation, and security of communication networks.
Joint work with Eran Simhon.