Abstract: Abstract: In cognitive radio networks, according to the robust and multiusers joint problem of multiuser admission control and power beamforming, L0-norm minimization is used to realize the joint optimization of multiuser admission control and power beamforming. Take the L0 norm of non-convexity and discontinuity into consideration, a joint optimization program based on a smooth similar of entropy function is proposed. Since the objective function is smooth differentiable and unimodal functions, the problem is transformed into the augmented Lagrange function, using the Armijo gradient search algorithm to obtain the optimal solution of the problem. Numerical results showed that compared with other joint optimization algorithm, the proposed algorithm has less the number of admitted uses when SINR is relatively low, but the proposed algorithm can reduce the transmission power consumption and improve the number of admitted uses when SINR is relatively high.

Key words: cognitive radio network, admission control, transmission power beamforming, the augmented Lagrange function, the Armijo gradient search algorithm

摘要： 摘 要: 在认知无线网络中，针对多用户鲁棒性的接入控制和发射功率波束形成的联合优化问题，利用L0-范数最小化来实现接入控制和发射功率波束的联合优化。考虑到L0-范数的非凸性及不连续性，提出了基于熵函数光滑近似的联合优化方案。由于光滑可微的目标函数为单峰函数，将问题变形为增广Lagrange函数利用Armijo梯度下降法得到问题的最优解。数值结果分析表明：相比其他的联合优化算法，当信干噪比较低时所提算法的接入量略有下降，但是当信干噪比较高时所提算法能显著降低发射功率并提高次用户的接入量。

关键词: 关键词: 认知无线网络, 接入控制, 发射功率波束形成, 增广Lagrange函数, Armijo梯度下降法