A hybrid game model based on reputation for spectrum allocation in wireless networks

With the dramatic development of communication technology, wireless network-based applications become more and more popular. Among them, network convergence is the typical scenario that interconnects various networks or migrates different data or services into a uniform system. Accordingly, spectrum resource allocation becomes an urgent problem, especially for convergent networks. In order to fully utilize the scarce spectrum resource and improve the QoS of network, we propose a spectrum allocation scheme by constructing the hybrid Game Model Based on Reputation (GMBR) in wireless network, which deals with multiple primary and secondary users coexisting and sharing the spectrum. First, we establish the static game model, and then, we evolve it as a multi-stage repeated game. On account of the users’ selfishness, the reputation mechanism is adopted to optimize the overall spectrum efficiency while keeping the participating incentives of the users. Finally, the stability condition of the dynamic action for GMBR is investigated. The analysis and simulation results show that GMBR has faster convergence speed than theoretical competitive equilibrium in the same QoS requirement, which means less energy and resource consumption.