针对多输入多输出(MIMO)雷达正交相位编码信号设计中已有方法所得信号集匹配滤波后旁瓣较高的问题,提出一种新的设计方法.首先将基于信号集相关旁瓣总能量的代价函数简化成低次的较容易求解的问题,然后将输入输出算法从一维扩展到多维,并针对极小化问题改进了其物面和频谱面的约束条件,利用改进后的输入输出算法求解该极小化问题,再针对基于信号集相关旁瓣峰值的代价函数,利用迭代码选择算法对第一步所得信号集进一步优化,最后得到相关性能较好的正交信号集.仿真结果表明,相比于模拟退火和序列二次规划等方法,该方法所得信号集的平均非周期自相关峰值旁瓣降低了3~4 dB,平均非周期互相关峰值降低了1~2 dB.
In the design of orthogonal phase coding signals for multi-input multi-output (MIMO) radar, a new method is presented to solve the problem that the signal sets obtained by the existing algorithms have high integrated sidelobes after matched filtering. An optimization cost function based on the total correlation sidelobes energy of the signal set is converted into a simpler quadratic optimization problem. An input-output algorithm is amended and extended to multidimensional case to improve the constraints of object domain and frequency domain, and the improved input-output algorithm is used to get a solution of the simplified optimization problem. Then, the results from the first step are further optimized by minimizing the correlation sidelobes by using the iterative code selection method, and the signal set with low correlation sidelobes is finally generated. Simulation results and comparisons with the simulated annealing algorithm and the sequential quadratic programming ( SQP ) algorithm, etc. autocorrelation sidelobe peak and the average crosscorrelation peak of the presented method decrease by about 3-4 dB and 1-2 dB, respectively.