一种精确的基于周期序列的信噪比估计算法  

Precise SNR Estimation Algorithm Based on Periodical Sequence

作　　者：余晟[1] 程云鹏[2] 李振海[1] 王乃超[1] 

机构地区：[1]解放军理工大学通信工程学院研究生3队,江苏南京210007 [2]解放军理工大学通信工程学院

出　　处：《军事通信技术》2013年第1期21-25,36,共6页Journal of Military Communications Technology

摘　　要：文中针对多径信道,提出了一种基于周期序列的信噪比估计算法,利用周期序列良好的自相关特性,精确地估计了接收信号幅度和噪声方差,分析并推导了两径信道下信噪比估计的通用表达式,给出了算法实现的具体过程。仿真结果表明该算法整体估计性能较好,特别适合于低信噪比条件下。与现有的二阶矩四阶矩(M2M4)和频域估计算法相比,在信噪比为-5dB时,估计偏差在1dB范围内的概率分别提高了21%和55.5%。For the multipath channel, a Signal Noise Ratio （SNR） estimation algorithm was presented based on periodical sequence. With the sound autocorrelation characteristic of periodical sequence, the signals＇ amplitudes and the noise＇s variance were precisely estimated, and a common expression of SNR estimation in two-path channel analyzed and deduced in detail. On the other hand, the specific implementation of the algorithm was also proposed. Simulation resuits show that this algorithm performs better than the 2-order and 4-order moments （M2M4） and frequency domain estimation algorithm, especially under the condition of low SNR. As the actual SNR is --5 dB, the probability of the estimation biases within 1 dB improves 21% and 55.5% respectively.

关 键 词：周期序列 信噪比估计 自相关特性 

分 类 号：TN92[电子电信—通信与信息系统]