| 摘要: |
| 对不平坦噪声基底的准确估计是宽带接收机信号检测的关键，传统的宽带检测多采用基于迭代形态学运算的静态噪底估计方法，不同时刻的估计结果之间互不相关，存在分辨率低、迭代耗时和时间连续性差等问题。针对上述问题，提出了一种动态噪底跟踪方法。首先建立带内各频点噪声基底的时变模型，然后将其作为系统状态引入卡尔曼滤波框架进行时间维度的噪底连续估计，同时结合频率维度的高斯平滑，实现了时间-频率两个维度的噪底跟踪。通过实采宽带信号验证，所提算法相较传统方法在噪底估计分辨率、连续性、计算效率等方面均有明显提升。 |
| 关键词: 宽带接收机 信号检测 静态噪底估计 动态噪底跟踪 卡尔曼滤波 高斯平滑 |
| DOI：10.20079/j.issn.1001-893x.220629007 |
|
| 基金项目: |
|
| Signal detection based on dynamic noise floor track in broadband receiver |
| LI Yu,LIANG Xianming,LONG Huimin |
| (Southwest China Institute of Electronic Technology,Chengdu 610036,China) |
| Abstract: |
| Estimating the undulating noise floor is the key to signal detection of broadband receiver.Existing methods generally employ static noise floor estimation that are based on iterative morphological operation,which means the estimated results of different time instance are independent.These methods suffer from poor resolution and continuity as well as huge computing cost.A novel noise floor estimation method is proposed to solve these problems.According to the method presented,the noise floor corresponding to every bin is modeled as time-varying state and introduced into Kalman filter scheme to achieve noise floor estimation in time dimension.Combined with the incorporation of Gaussian smoothing in frequency dimension,noise floor tracking in both time and frequency dimension is finally realized.Experimental results demonstrate that the proposed method has better performance in resolution,continuity and computational efficiency than the traditional morphological method. |
| Key words: broadband receiver signal detection static noise floor estimation dynamic noise floor tracking Kalman filter Gaussian smoothing |
