Spectrum sensing is one of the critical tasks in a cognitive radio system that allows a secondary user to use the spectrum while the primary user does not use it. The energy detection (ED) sensing is one of the most common techniques to identify the unused portions in the spectrum bands. In ED, threshold plays a vital role in signal detection, and noise is one of the significant factors in threshold calculation. However, ED efficiency is degraded by the noise uncertainty phenomenon caused by the random changes in noise level. The adverse effects of noise uncertainty are reduced by changing its detection threshold dynamically to the noise circumstances encountered during each sensing period. In the proposed method, received random samples are arranged in M blocks, applied strong Pearson correlation to separate and estimate the variance from the noise samples. The enhanced dynamic noise variance‐based energy sensing is implemented in GNU radio processing blocks and tested on industrial, scientific, medical (ISM) 2.4 GHz frequency bands by using national instrument universal software radio peripheral (NI USRP‐2932) device. The experimental results of proposed energy detction mechanism are compared with existing sensing techniques.

中文翻译：

---

讨论：在Wi-Fi频段使用USRP的基于增强型动态噪声变化的EneRgy感知无线电认知无线电

频谱感测是认知无线电系统中的关键任务之一，它允许次要用户使用频谱而主要用户不使用频谱。能量检测（ED）感测是识别频谱带中未使用部分的最常用技术之一。在ED中，阈值在信号检测中起着至关重要的作用，而噪声是阈值计算中的重要因素之一。然而，由于噪声水平的随机变化而引起的噪声不确定性现象使ED效率降低。通过动态更改其检测阈值以适应每个感测周期遇到的噪声情况，可以降低噪声不确定性的不利影响。在提出的方法中，接收到的随机样本按M排列块，应用强Pearson相关来分离和估计噪声样本的方差。在GNU无线电处理模块中实现了增强的基于动态噪声方差的能量传感，并使用国家仪器通用软件无线电外围设备（NI USRP‐2932）在工业，科学，医学（ISM）2.4 GHz频段上进行了测试。将提出的能量检测机制的实验结果与现有的传感技术进行了比较。