CAJ | 학술논문

针对现有方法分选准确率不高和对噪声敏感的问题，应用信息熵特征实现了低信噪比条件下未知复杂雷达信号的高准确率分选。对接收到的雷达信号首先求得维格纳分布空间和双谱空间，然后对该空间进行子空间划分，在子空间分布概率下，求取雷达信号的信息熵特征构成二维特征向量，该特征向量反映了信号在这两种空间下的能量分布。最后对8种雷达辐射源信号进行了特征提取和分类仿真实验。实验结果表明所提取的特征参数在宽信噪比范围内均具有很好的分离性和稳定性，验证了该方法的有效性。
침대현유방법분선준학솔불고화대조성민감적문제，응용신식적특정실현료저신조비조건하미지복잡뢰체신호적고준학솔분선。대접수도적뢰체신호수선구득유격납분포공간화쌍보공간，연후대해공간진행자공간화분，재자공간분포개솔하，구취뢰체신호적신식적특정구성이유특정향량，해특정향량반영료신호재저량충공간하적능량분포。최후대8충뢰체복사원신호진행료특정제취화분류방진실험。실험결과표명소제취적특정삼수재관신조비범위내균구유흔호적분리성화은정성，험증료해방법적유효성。
Focusing on the problems that the current radar signal sorting method is of low accuracy rate and high sensitivity to noise, this paper implements the high accuracy sorting of unknown complex radar signals under the condition of low signal-to-noise ratio, using the information entropy features. At first, Wigner distribution space and bi-spectrum space are obtained to the received radar signals, next, those spaces are divided into sub-spaces, and then, under the distribution probability of sub-space, the information entropy features of radar signals are found to form the two dimension eigenvectors, which reflect the signals’energy distribution under those two spaces. And finally, the characteristics extraction and simulation experiment on classification are performed to eight types of radar emitter signals. Experimental results show that the extracted characteristic parameters are of excellent separability and stability in a wide range of SNR, thus, proving the feasibility of this proposed method.