Abstarct: Usage of smart grids has witnessed an increase in the recent years, including its usage in renewable energies such as solar and wind energies. Optimal usage of these new energies demands exchange of data. In contrast to the previous energy grids, consumers, producers, and other possible nods involved in smart grids exchange data with each other. Various methods exist for communication among the nods in a smart grid. As it is discussed in this work, the older wired method and the wireless methods suffer shortcomings for their application in smart grids. Cognitive radio, with its reliance on configurability of its parameters, may be the best viable option for smart grids. The present dissertation has attempted to apply game theory techniques in implementing a new and optimized system model for resource allocation in smart grids. We developed and compared four algorithms for resource allocation.