基于二次约束二次规划的窗函数设计方法

(1.中国科学院 空天信息创新研究院,北京 100094; 2.中国科学院大学 电子电气与通信工程学院,北京 100049)

雷达信号处理; 凸优化; 多目标优化; 窗函数设计; 成像质量

A window function design method via QCQP approach
LIU Jin1,2,SONG Hong-jun1

(1.Aerospace Information Research Institute,Chinese Academy of Sciences,Beijing 100094,China; 2.School of Electronic Electrical and Communication Engineering,University of Chinese Academy of Sciences,Beijing 100049,China)

radar signal processing; convex optimization; multi-objective optimization; window function design; imaging quality

DOI: 10.13800/j.cnki.xakjdxxb.2020.0311

备注

为降低高功率雷达系统中脉冲压缩后输出的高峰值旁瓣进而提高成像质量,通常的方法是在接收端进行加窗处理,在信噪比损失、积分旁瓣和主瓣宽度上做出一定权衡。为了进一步提升窗函数的成像性能,以不同的目标函数和约束条件构建能够高效准确求解的凸优化模型来设计优化窗函数。建立了以优化峰值旁瓣为目标的二次约束二次规划的凸优化模型,并给出了优化积分旁瓣能量的凸优化模型。通过线性加权求和,对峰值旁瓣、信噪比损失和积分旁瓣能量的线性组合得到的结果作为目标函数建立了多目标凸优化模型。通过对凸优化问题求解,得到的优化窗函数在峰值旁瓣、积分旁瓣、信噪比损失和主瓣宽度上均优于或不劣于在雷达系统中常用的性能优良的Taylor窗函数。仿真结果表明,优化窗函数能够比Taylor窗函数提供更好的性能。

To lower the high peak sidelobe ratio(PSLR)generated from the output of pulse compression in the high power radar system and thus improve the imaging quality,a common method is to apply window function at the reception side,which can make a trade off among the signal-to-noise ratio(SNR)loss,integrated sidelobe ratio(ISLR)and mainlobe width.In order to further improve the imaging performance of the window function,a convex optimization modelwas proposed to design the optimized window function,and this model was constructed by usingvarious objective functions and constraints which can be solved efficiently and effectively.First,a quadratically constrained quadratic program(QCQP)convex optimization model was built to optimize the PSLR,and based on this model,a convex optimization model aiming at optimizing the integrated sidelobe energy was proposed.By linear weighted summation,a multi-objective convex optimization model was proposed aiming at optimizing the linear combination of PSLR,SNR loss and the integrated sidelobe energy.By solving the convex optimization problem,the optimized window function can be obtained,which can provide better or no worse PSLR,ISLR,SNR loss and mainlobe width than the well performed Taylor window.The results show that the optimized window functions can offer better performance than Taylor window function.