Game theory based dynamic spectrum sharing for next-generation mobile networks

Next-generation mobile networks need to ensure flexible and effcient utilisation of the radio spectrum to keep pace with the rapid growth of mobile services and applications in recent years. This thesis presents novel game-theoretic frameworks for enabling dynamic spectrum sharing to achieve spectral effciency in next-generation mobile networks, and proposes novel ideas for dynamic allocation of spectrum resources among participating networks, while ensuring reliable quality of service (QoS) for all users through appropriate sharing rules, protocols and architectures. The spectrum-sharing schemes proposed in this thesis are categorised as power-control games for femtocell-based networks and games based on Licensed Shared Access (LSA). The main contributions of this thesis include formulating equal-priority and multi-priority non-cooperative power-control games for femtocell-based networks to achieve interference mitigation such that the QoS of the users served by the femtocells is not compromised. During the game, each femtocell, serving either the primary users or the secondary users of the spectrum, dynamically adjusts its transmit power until the transmit power is stabilised, adapting to its interferenc emeasurement and traffc. It is shown that the existence and uniqueness of the Nash equilibrium of the non-cooperative game can be achieved by carefully selecting the game parameters, which are designed to determine the priority status of the femtocells. Furthermore, the real-time signalling overhead is minimised between the networks through a novel dual-mode solution. Another key contribution of this thesis is the application of game-theoretic principles for dynamic spectrum sharing using LSA. A two-layer LSA-based evolutionary game is proposed which ensures demand-driven allocation of spectrum resources to LSA licensees. The scheme also incorporates dynamic price-adjustment strategies adopted by incumbents to ensure an improved total gain for the incumbents.The stability of the proposed evolutionary algorithm is proved using Lyapunov stability criteria. Furthermore, a practical scenario of LSA-based spectrum sharing is considered, and the LSA architecture is modified to enable mobile network operators acting as domestic licensees to provide an enhanced QoS in border areas, for minimising the effects of cross-border interference.