摘要
AVC算法是一种适用于拉普拉斯噪声环境的常用频谱感知算法,但该算法并未充分平滑拉普拉斯噪声中的"尖峰",导致算法的检测性能不佳。针对此,提出一种改进的AVC频谱感知算法,其原理是对接收信号绝对值做开根号处理,并累加处理结果,作为检验统计量,进而判决是否存在主用户,实现频谱感知。此外,利用中心极限定理推导了所提算法检验统计量在主用户不存在时的概率密度曲线,从而给出理论判决门限。仿真表明,所提算法的检测性能分别优于AVC感知算法和拉普拉斯噪声下的能量检测算法大约1 d B和4 d B。
Absolute value cumulating （AVC） algorithm is a common spectrum sensing method in Laplacian noise （LED） surroundings, however, the ＇spikes or outliers＇ in Laplacian noise can＇t be fully smoothed, which results in bad detection performance. Aiming at this problem, an modified AVC spectrum sensing algorithm was proposed. The principle was to do the processing of the absolute value of the received signal and the processing result was accumulated as the test statistic to determine whether there was the main user and realize the spectrum sensing. In addition, the central limit theorem was used to deduce the probability density curve of the proposed test statistic in the absence of the primary user, and the theoretical decision threshold was given. Numerical results show that the proposed algorithm outperforms the AVC sensing algorithm and energy detection algorithm with Laplacian noise about 1 dB and 4 dB respectively.
出处
《电信科学》
北大核心
2017年第2期98-103,共6页
Telecommunications Science
基金
江苏风力发电工程技术中心开放基金资助项目(No.ZK16-03-13)
南京工业职业技术学院学校科研基金资助项目(No.YK16-02-01)~~
关键词
频谱感知
拉普拉斯噪声
中心极限定理
理论判决门限
spectrum sensing, Laplacian noise, central limit theorem, theoretical detection threshold
