论文部分内容阅读
研究网络环境下具有随机丢包的自回归滑动平均(ARMA)信号的估计问题,其中丢包现象通过一个满足Bernoulli分布的随机变量描述.通过ARMA模型与状态空间模型的转化,将具有丢包的ARMA信号估计问题转化为具有丢包的状态空间模型的状态和白噪声估计问题.利用射影理论分别给出线性最小方差最优线性状态估值器和白噪声估值器,进而获得ARMA信号估值器.仿真结果表明,当存在数据丢失时,所提出的算法与以往基于完整数据的最优估计算法相比具有最优性和有效性.
This paper studies the estimation problem of autoregressive moving average (ARMA) signals with random packet loss in network environment, in which the packet loss phenomenon is described by a random variable that satisfies the Bernoulli distribution. By transforming ARMA model and state space model, ARMA signal estimation problem is transformed into the state of the state space model with packet loss and the white noise estimation problem.The linear minimum variance optimal linear state estimator and the white noise estimator are respectively given by the projective theory and the ARMA signal estimate The simulation results show that when there is data loss, the proposed algorithm is more optimal and effective than the best one based on the complete data.