Combining Cooperative With Non-Cooperative Game Theory to Model Wi-Fi Congestion in Apartment Blocks

The unlicensed spectrum utilized by Wi-Fi can be regarded as an economic commons in many deployments. Operators of Wi-Fi-enabled devices are usually non-cooperative, vying for spectral resources when in close range to each other, typically adopting a strategy of maximizing their transmission power. With an ever-growing number of wireless devices, this will ultimately lead to depletion of the spectrum, unless players collaborate. Previous studies used cooperative game theory to explore various collaboration strategies, enabled by the presence of some central authority or controller that executes an agreed-upon interference mitigation policy. However, the regulatory nature of unlicensed spectrum dictates that players cannot be forced into such collaboration. Most deployments therefore involve a mix of cooperative and non-cooperative players. In this paper, we propose a new way of modeling use cases involving a central authority or controller by combining non-cooperative and cooperative game theory. Our model uses the non-cooperative concept of Nash equilibriums as well as the cooperative concept of Nash bargaining. To the best of our knowledge, this paper is the first to propose a hybrid non-cooperative and cooperative game theoretic model for communication networks that offers the players the opportunity to strategize between non-cooperative and cooperative natures. It is successfully applied to the case of a densely-populated apartment block. The results show that, if only a subset of players joins the collaboration, most of the remaining non-joining players may obtain an SINR that is worse than what they would have obtained in the fully non-cooperative scenario; they are punished for their uncooperative behavior.