Information theory and game theory

From Webresearch

Jump to: navigation, search

Game theory is a natural framework to study competitive environments. Generally, a game involves a set of players with potentially conflicting objectives. Each player can choose among a set of actions according to a strategy which would interactively affect the utilities of the other players. These interactions, either cooperative or noncooperative, will evolve until a state of equilibrium is reached. Depending on the setting of the game, various notions of equilibria arise; Nash, saddle-point, correlated, and Stackelberg equilibria. Game theory can be categorized into cooperative and noncooperative, dynamic and static, complete and incomplete information, perfect and imperfect information, etc.

Recently, there have been an increasing interest in applying game-theoretical concepts to investigate topics in communications, signal processing, security, biology, control, and information theory. Based on the nature of the problem, game theory can be utilized for performance analysis or to design system parameters in order to reach a desired performance. Specific examples include resource allocation, evolutionary biology, social network games, security and evolution of trust, mechanism design, etc. Long standing problems in information theory have recently been reexamined using game-theoretical tools leading to additional insights and introducing new perspectives.

Personal tools