2016/08/25 by Mehrdad Khaledi, Khaledi, Mehrdad, Alhussein A. Abouzeid +1
Decision Sciences · Engineering · #Advanced Bandit Algorithms Research #Auction Theory and Applications #Computer Science and Game Theory (cs.GT) #FOS: Computer and information sciences #Networking and Internet Architecture (cs.NI) #Smart Grid Energy Management
paper · pdf · doi:10.48550/arxiv.1608.07357
openalex publication_date 2016/08/25 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28
Small operators who take part in secondary wireless spectrum markets\ntypically have strict budget limits. In this paper, we study the bidding\nproblem of a budget constrained operator in repeated secondary spectrum\nauctions. In existing truthful auctions, truthful bidding is the optimal\nstrategy of a bidder. However, budget limits impact bidding behaviors and make\nbidding decisions complicated, since bidders may behave differently to avoid\nrunning out of money. We formulate the problem as a dynamic auction game\nbetween operators, where knowledge of other operators is limited due to the\ndistributed nature of wireless networks/markets. We first present a Markov\nDecision Process (MDP) formulation of the problem and characterize the optimal\nbidding strategy of an operator, provided that opponents' bids are i.i.d. Next,\nwe generalize the formulation to a Markov game that, in conjunction with\nmodel-free reinforcement learning approaches, enables an operator to make\ninferences about its opponents based on local observations. Finally, we present\na fully distributed learning-based bidding algorithm which relies only on local\ninformation. Our numerical results show that our proposed learning-based\nbidding results in a better utility than truthful bidding.\n