Inter-operator spectrum sharing for cellular networks using game theory

In this paper, we present a game theoretical framework for DSA (Dynamic Spectrum Access) in cellular networks. We model and analyze the interaction between cellular operators with packet services, in a spectrum sharing context. We present inter-operator DSA algorithms based on game theory. A two-players non-zero sum game is formulated, where the operators are the players. We define a utility function, for the operator that takes: (1) the users throughput, (2) the spectrum price, and (3) the blocking probability into consideration. We present two system models: a) a centralized model where a DSA algorithm, for the global welfare in terms of the operators rewards, is inspired by the Pareto optimality concept. b) a distributed model, where a DSA algorithm is based on Nash equilibria concept. The convergence to NE in the distributed model is analyzed. The rewards of the operators in the centralized DSA algorithm are compared with those in the FSA (Fixed Spectrum Access) situation. The obtained rewards using the centralized DSA algorithm significantly exceed the FSA rewards. The obtained blocking probabilities are shown not to exceed the target value.