中文摘要：

针对归一化最小均方（NLMS）算法步长调整精度不高的问题，将约束稳定性最小均方（CS-LMS）算法应用于智能天线的波束形成问题。通过引入后验误差的稳定性约束条件，构建适合波束形成问题的优化目标，并给出了其解的计算复杂度、收敛速度和稳态误差的理论分析。计算机仿真结果证实，CS-LMS波束形成算法比NLMS算法具有更加优良的抗干扰性能，较快的收敛速度，良好的系统跟踪能力和较小的稳态误差，尤其是当输入信号发生中断或突变时，算法依然工作得很好。

英文摘要：

Because the adjustment of step size in normalized least mean square （NLMS） algorithm is not accurate enough,the constraint-stability least-mean-square（CS-LMS） algorithm is applied for beamforming of smart antenna.By introducing a constraint-stability condition to the posterior errors,a new optimization func-tion suitable for beamforming problem is set up and the computational complexity,convergence speed and steady state error of the solution are also analyzed theoretically .Computer simulation results show that the CS-LMS algorithm outperforms the NLMS algorithm in terms of anti-interference performance,convergence rate,system tracking ability and steady state error .The CS-LMS can also work well when the interruption or change of input signal occurs suddenly .