Abstract
In economics and algorithmic game theory, the Facility Location Game is a kind of competitive game in which social planners select locations to place facilities in order to maximize the social benefits for serving agents. Given that agents have private information, asymmetric games are considered strategyproof (or truthful), if it is a weaklydominant strategy for every agent to disclose his private information, i.e., no matter what the others do, you should perform best or at least not worse by being truthful. In this thesis, we present three projects, all of which are concerned with strategyproof mechanism design in the facility location game. The first project is strategic learning approach for deployingUAV-provided wireless services. Unmanned Aerial Vehicle (UAV) has emerged as a promising technique to provide wireless services to a group of mobile users simultaneously. We aim to address a challenging issue that each user is selfish and may misreport his location or preference for changing the optimal UAV location to be close to himself. By considering the UAV as a new facility in the facility location game, we study how to determine the final location of aUAV in the 3D space, by ensuring all selfish users’ truthfulness in reporting their locations. We design strategyproof mechanisms to cover all cases where the UAV is favorable, obnoxious, and also extend all games’ design to include multiple UAVs. The second project focuses on the two-facility location game with minimum distance requirement. We study the mechanism design problem of a social planner for locating two facilities on a line interval, where a set of strategic agents report their locations and a mechanism determines the locations of the two facilities. Unlike prior work, we consider the requirement of the minimum distance between the two facilities. On the one hand, the two facilities are either heterogeneous, or homogeneous; on the other hand, each one of the two facilities can be favorable, obnoxious, or indifferent for any agent. We propose a family of strategyproof mechanisms with provable approximation ratios, which are suitable for all possible scenarios. Strategyproof mechanism design for activity scheduling is the last project. We extend the facility point in the facility location game to be an interval and study a novel activity scheduling game to schedule an activity in the time domain based on all agents’ time reports. The activity lasts for a time period, and each agent wants his private time to be within the activity duration or at least as close as possible. Our objective is to choose the activity starting time so that the mechanisms are strategyproof. We design a strategyproof mechanism outputting an optimal solution. Besides, we also study the obnoxious activity scheduling game, and consider the cost/utility as the characteristic functions to design various strategyproof mechanisms.