Active Cooperation-aware Spectrum Sharing And Power Allocation In Cognitive Radio Networks

Cooperative cognitive radio network(CCRN), combining cognitive radio technology with cooperative communication technology, can effectively improve spectrum resource utilization. However, in the traditional CCRN, spectrum sharing is lack of flexibility, due to its inherent dynamic heterogeneous characteristics, and primary user(PU)’s priority on spectrum. For this reason, a novel active cooperation frameworks between PU and secondary users(SU) is proposed, which can significantly improve the utilization of spectrum resources and increasing data throughput and coverage of wireless networks. The introduction of active cooperation frameworks may bring some new challenges to the resources allocation in CCRN. The optimization of spectrum sharing and power allocation in the framework is also considered, pursuing the solution of improving spectrum utilization and energy efficience.Firstly, we focus on spectrum resources sharing in CCRN and present a novel active cooperation based collaborative incentive strategy, where the primary user is aware of the exist of second users. Specific to spectrum property features and nodes’ selfish nature of cognitive radio network, we model the spectrum sharing between primary users and secondary users as stackelberg leader-follower game, based on analysising the benefits and collaboration relationship each other. In the proposed framework, the PU can lease a portion of its spectrum to the SUs in order to stimulate them to actively join the primary user cooperative transmission, when its QoS is guaranteed. SUs cooperating with primary transmissions have the right to decide their payment, aiming at maximizing its own utility. The backward induction method is applied to analysis and prove the existence and uniqueness of the stackelberg equilibrium solution. In addition, the fairness and service robustness (potentially continuous) are also investigated.Secondly, we further consider the complex dynamic cognitive network scenarios with multiple primary users and secondary users, and address the SUs’optimum power allocation proplem, propose a new way to encourage primary users to lease their spectrum. The second-price auction is applied to achieve optimal channel assignment strategy. Allowing for the difference of link quality and communication requirement of users, robust distributed reinforcement learning based power allocation approach is proposed, relieving energy constraint effect and improving the energy efficiency. The reinforcement learning mechanism is developed to allow the users to revise their bids and to increase their rewards, eliminating the impact on network performance of the channel state information imperfectness and the environment uncertainty.Finally, the performance evaluations of the proposed active cooperation-aware spectrum sharing and power allocation are taken via MATLAB simulation. The performance results show the significant impact of proposed algorithm in both improving spectrum utilization and energy consumption.