Bayesian estimation of mean transmission loss along multiple paths with randomly scattered signals
Bayesian estimation of mean transmission loss along multiple paths with randomly scattered signals
复制标题
具有随机散射信号的多条路径平均传输损耗的贝叶斯估计
DOI：
--
复制
发表时间：
2018
影响因子：
2.4
通讯作者：
V. Ostashev
中科院分区：
文献类型：
--
作者：
D. Wilson;C. Pettit;V. Ostashev
考虑了一个未知功率源向多个接收点传输的一般问题。信号在每条传输路径上随机分散，例如湍流、森林或建筑物。在每个接收器位置，收集信号功率的一个或多个观测值。假设接收信号的概率密度函数（pdf）的函数形式（例如，强散射的指数pdf或弱散射的伽马pdf）是基于对散射过程的理解而已知的，尽管从源到接收器的平均传输损耗（TLs）是不确定的。从接收机的信号观测中，我们希望估计源的平均功率或每条路径的平均TLs。提出了该问题的贝叶斯公式，该公式对未知源功率使用逆先验，对未知TLs使用矩阵逆先验，对随机散射信号使用Wishart分布。由于伽马先验是Wishart分布的共轭，因此可以导出解析解，以便在进行新信号观测时更新源功率或TL估计。考虑了一个未知功率源向多个接收点传输的一般问题。信号在每条传输路径上随机分散，例如湍流、森林或建筑物。在每个接收器位置，收集信号功率的一个或多个观测值。假设接收信号的概率密度函数（pdf）的函数形式（例如，强散射的指数pdf或弱散射的伽马pdf）是基于对散射过程的理解而已知的，尽管从源到接收器的平均传输损耗（TLs）是不确定的。从接收机的信号观测中，我们希望估计源的平均功率或每条路径的平均TLs。提出了该问题的贝叶斯公式，该公式对未知源功率使用逆先验，对未知TLs使用矩阵逆先验，对随机散射信号使用Wishart分布。自从…
A general problem is considered in which a source of unknown power transmits to multiple receiver locations. The signal is randomly scattered along each transmission path, for example by turbulence, a forest, or buildings. At each receiver location, one or more observations of the signal power are collected. It is assumed that the functional form of the probability density function (pdf) of the received signals (e.g., an exponential pdf for strong scattering, or a gamma pdf for weak scattering) is known based on an understanding of the scattering process, although the mean transmission losses (TLs) from the source to receivers are uncertain. From signal observations at the receivers, we wish to estimate the mean power of the source or the mean TLs for each path. A Bayesian formulation for this problem is presented, which utilizes an inverse gamma prior for the unknown source power or a matrix inverse gamma prior for the unknown TLs, and a Wishart distribution for the randomly scattered signals. Since the gamma priors are conjugates of the Wishart distribution, analytical solutions can be derived for updating the source power or TL estimates as new signal observations are made.A general problem is considered in which a source of unknown power transmits to multiple receiver locations. The signal is randomly scattered along each transmission path, for example by turbulence, a forest, or buildings. At each receiver location, one or more observations of the signal power are collected. It is assumed that the functional form of the probability density function (pdf) of the received signals (e.g., an exponential pdf for strong scattering, or a gamma pdf for weak scattering) is known based on an understanding of the scattering process, although the mean transmission losses (TLs) from the source to receivers are uncertain. From signal observations at the receivers, we wish to estimate the mean power of the source or the mean TLs for each path. A Bayesian formulation for this problem is presented, which utilizes an inverse gamma prior for the unknown source power or a matrix inverse gamma prior for the unknown TLs, and a Wishart distribution for the randomly scattered signals. Since the ...