高斯尺度混合大气噪声模型的参数估计

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。

Parameters Estimation for Gaussian Scale Mixture Atmospheric Noise

Parameters estimation of the non-Gaussian atmospheric noise has an important significance to the signals optimal detection in very low frequency（ VLF） and extremely low frequency（ ELF） communica-tion. A model based on Gaussian scale mixture（GSM） distribution with inverse-Gaussian is proposed. The characteristic of parameters of GSM distribution with inverse-Gaussian is studied. Markov chain Monte Carlo（ MCMC） method is designed to estimate parameters. This method, based on Bayesian hierarchical model,updates the parameters through Gibbs sampler and M-H algorithm. The experimental results show that the proposed model performs well. The MCMC method is of good iterative efficiency and precision,and can be excellently applied in practice.