论文部分内容阅读
非高斯大气噪声的参数估计对甚低频、超低频信号的最佳接收有重要意义。对大气噪声采用基于逆高斯分布的高斯尺度混合分布模型建模,研究了基于逆高斯分布的高斯尺度混合分布模型参数的性质,设计了高斯尺度混合大气噪声模型参数的马尔可夫链蒙特卡罗(MCMC)算法。算法在贝叶斯层次模型下,采用Gibbs抽样和M-H抽样更新参数。仿真结果表明,该模型对大气噪声有很好的适用性,MCMC算法迭代效率和精度高,具有实际的应用价值。
Parameter estimation of non-Gaussian atmospheric noise is of great importance for the optimal reception of very low frequency and very low frequency signals. Gaussian scale mixed distribution model based on inverse Gaussian distribution was used to model atmospheric noise. The properties of Gaussian scale mixed distribution model parameters based on inverse Gaussian distribution were studied. The Markov chain Monte Carlo (MCMC) algorithm. The algorithm uses Gibbs sampling and M-H sampling to update the parameters under the Bayesian level model. The simulation results show that the model has good applicability to atmospheric noise. The MCMC algorithm has high iterative efficiency and high precision, and has practical value.