信息融合WIENER信号滤波器

利用现代时间序列分析方法,基于ARMA新息模型和增广的状态空间模型,提出了两传感器最优信息融合WIENER信号滤波器,给出了计算局部滤波器误差方差和局部滤波器之间协方差的计算公式,它们可被用于计算最优加权系数．同单传感器情形相比,可提高滤波器的精度．一个仿真例子说明其有效性．

Using self-location to calibrate the errors of observer positions for source localization

The uncertainty of observers＇ positions can lead to significantly degrading in source localization accuracy. This pa-per proposes a method of using self-location for calibrating the positions of observer stations in source localization to reduce the errors of the observer positions and improve the accuracy of the source localization. The relative distance measurements of the two coordinative observers are used for the linear minimum mean square error （LMMSE） estimator. The results of computer si-mulations prove the feasibility and effectiveness of the proposed method. With the general estimation errors of observers＇ positions, the MSE of the source localization with self-location calibration, which is significantly lower than that without self-location calibra-tion, is approximating to the Cramer-Rao lower bound （CRLB）.

OFDM电力线通信系统的时域信道估计方法

针对G3-PLC协议,提出了两种改进的时域信道估计方法。为达到时域降噪效果,采用了预置电力线信道最大时延的方法,但这使得相关的时频变换矩阵不满秩,无法直接采用最小二乘准则估计其时域特性。因此,文中提出一种基于修正变换矩阵的时域LS算法,解决了矩阵不可逆问题。同时进一步提出了一种无需信道统计特性的时域线性最小均方误差算法,大幅降低了复杂度。仿真结果表明,修正的时域LS算法和简化的时域LMMSE算法均性能良好。