Games Theory and Software Defined Radios

The use of game theory in software designed radio networks is studied using OPNET. In order to model the spectrum usage, radio frequency interference avoidance, and distributed radio resource management, the behavior of software defined radios are predicted using game theory in an analytic mathematical framework. The wireless networks consist of devices that dynamically reconfigure themselves to respond to any air-interface or data format. Interfering radios however affect the adaptation schemes used on a link by link basis. Cognitive radio is an enhancement on traditional software radio design that can help establish a design model. Cognitive radios employ a cognition cycle to alter their actions in response to changes in the environment through the use of state machines. A smart network presents particularly difficult challenges to the analysis of radio resource management, as changes that one node makes may influence the decisions that other nodes make, so network planning remains a difficult task. The radios are the players in the game; the strategies in the game are based on the transmitter choosing the value of the adaptive link characteristic that maximizes the spectral efficiency, while meeting a BER constraint. The goal of the game is to maximize your winnings. Here the winnings would be for a radio to accrue the most bandwidth, or in our case, to achieve a particular performance target. Additionally, these approaches rest upon the assumption of higher-order rationality, i.e. the ability of a node to independently and recursively analyze a best response to other network nodes OPNET´s pipeline processing can provide a solution path unavailable with other tools. We use OPNET to model changes in waveform appearance due to the interfering effects of neighboring communication links. In order to successfully model these networks, it will be necessary to determine if the network will eventually reach a steady state. With a game theoretic analysis, these network s- - teady states can be identified from the Nash Equilibriums of its associated game. In other traditional approaches this is not possible and demonstrates the more rapid convergence of the game results to modeling and simulation