학술논문

Asymmetric Battlefield Uncertainty in General Lotto Games
Document Type
Periodical
Source
IEEE Control Systems Letters IEEE Control Syst. Lett. Control Systems Letters, IEEE. 6:2822-2827 2022
Subject
Robotics and Control Systems
Computing and Processing
Components, Circuits, Devices and Systems
Games
Resource management
Uncertainty
Task analysis
Security
Finite element analysis
Bayes methods
Game theory
agents-based systems
Language
ISSN
2475-1456
Abstract
How to strategically allocate resources against opponents is a central component in adversarial decision-making. Moreover, informational asymmetries often exist between competitors and can significantly impact outcomes. In this letter, we study General Lotto games under varying levels of informational asymmetry between two players. The General Lotto game is a popular model of competitive resource allocation, in which opposing players compete over a set of valuable battlefields. We consider information here as knowledge of the battlefield values, which are randomly drawn from a finite distribution. First, we consider a class of games in which one player observes all realized values, and the opponent does not observe any. We then consider scenarios where the opponent observes a subset of the realized values. In both settings, we completely characterize equilibrium payoffs and strategies. We then provide improvement factors that result from acquiring information, demonstrating that information can significantly improve one’s competitive performance.