中文摘要：

为了解决认知网络功率控制问题，根据非合作博弈论和干扰温度，设计出一种新型效用函数。该函数综合考虑授权用户和认知用户功率分配的要求，满足用户需求。并通过此效用函数提出一个基于非合作博弈论的新颖功率控制算法，分析了该算法的收敛性、纳什均衡解的存在性和唯一性。仿真结果表明，该算法在保证认知用户需求前提下，传输功率更加合理，体现出算法的有效性和公平性。与参考算法相比，节省了30．12％的功耗，并且具有较好的抗背景噪声。最后，与Koskie—Gajic（KG）算法相比，该算法具有较大的网络容量。

英文摘要：

To resolve the problem that the power control of cognitive network, according to non-cooperative game theory, and interference temperature, this paper proposed a new utility function considering the power distribution requirements of primary users（PUs） and secondary users（SUs）. It deduced a novel power control algorithm based on non-cooperative game theory. It also proved its convergence and the existent and unique of the Nash equilibrium. Simulation shows that this algorithm is effec- tiveness and fairness which can stratify the SINR demands and make secondary users have proper transmission power levels. Compared with the referenced algorithm, it not only can save 30.12% power consumption and achieve better anti-noise effect, but also have better capacity relative to Koskie-Gajie （KG） algorithm.