基于GNURadio分段差分双阈值的宽带信号检测技术研究与实现

Research and Implementation of Segmented Differential Dual Threshold Wideband Signal Detection Technology based on GNURadio

为了解决信道色噪声造成噪底不平坦条件下的宽带信号检测问题,提出将宽带信号分段采集,对窄带功率谱估计值进行差分形成对称阈值,然后根据窄带双阈值检测信号对整个带宽信号进行检测。该方法不要任何的先验信息,分段后各自计算阈值进行检测,解决了窄带间信号切分后信号个数错误、不平坦噪底单阈值误判等问题,实现了对宽带信号的检测。此外,设计了基于GNURadio和USRP软件无线电平台的宽带信号检测器,并通过系统验证了检测器的性能。理论和实验仿真表明:算法能够在宽频带内有效克服色噪声,对噪声不确定度具有鲁棒性,且由于运算复杂度低,可以对宽带信号进行实时检测。

In order to solve the problem of wideband signal detection under the condition of noise floor unevenness caused by channel color noise,segmentation acquisition of wideband signals is proposed.The narrowband power spectrum estimates are differentially formed to make a symmetric threshold,and the entire bandwidth signal is detected on the basis of narrowband dual threshold detection signal.This method requires no prior information,and each step is calculated after the segmentation.This solves the problem that the number of narrow-band signals is erroneously divided,and the problem that the single-threshold value of the uneven noise is misjudged,and thus the detection of the wideband signal is realized.A wideband signal detector based on the GNURadio and USRP software radio platforms is designed and the performance of the detector is verified by the system.Theoretical and experimental simulations indicate that this algorithm could effectively overcome color noise in a wide frequency band and is robust to noise uncertainty.Wideband signals can be detected in real time due to low computational complexity.