面向子带分解的非均匀抽取均衡算法

摘要:

宽带数字阵列雷达通道失配严重的情况下，带宽分割法相比逆傅里叶变换法均衡效果更好，但其在不完全分解时重建误差较大且占用硬件资源过多。针对上述问题，该文提出一种延长带宽非均匀抽取均衡算法。首先，设定需要划分的子带数目，根据阻带截止频率的位置合并不参与均衡的子带；然后，根据待均衡子带的频率特性对其进行非均匀抽取；最后，对抽取后的待均衡子带实现均衡、插值和合并操作。仿真结果表明，在均衡子带个数和重建误差相等的情况下，相比带宽分割法，提出方法的幅度改善因子和相位改善因子分别提高5.39 dB和7.48 dB，降低了运算复杂度和硬件资源占用率。

Abstract:

In the case of serious channel mismatch of wideband digital array radar, the band-partition method has a better equalization effect than the inverse Fourier transform method. However, the reconstruction error is significant, and the hardware resources required are excessive when the decomposition is incomplete. Aiming at the above problems, an extended bandwidth non-uniform extraction equalization algorithm is proposed in this paper. Firstly, the number of sub-bands to be divided is set, and sub-bands that do not participate in equalization are merged according to the position of stopband cutoff frequency. Then, non-uniform decimation is performed according to the frequency characteristics of the sub-bands to be equalized. Finally, the operations of equalization, interpolation, and merging are implemented for the sub-bands to be equalized after extraction.? The simulation results show that the amplitude improvement factor and phase improvement factor of the proposed method are 5.39 dB and 7.48 dB higher than that of the band-partition method when the number of equalized sub-bands and reconstruction error are equal, while reducing computational complexity and hardware resource usage.