论文部分内容阅读
In this paper, new Cramér-Rao lower bounds (CRB) of the estimates of frequencies, two-dimensional arrival angles and polarization parameters of multiple incident signals are derived for a polarization sensitive array. The incident sources have distinct carrier-frequencies, in contrast to the modeling of all sources to be at the same known carrier-frequency, which has been investigated in the existing research literature on the Cramér-Rao bounds (CRB) for polarization sensitive direction finding. The derived CRBs are compact closed-form expressions and applicable to an arbitrary array geometry. Numerical examples and analysis of some special cases provide insights into the fact that the estimation accuracy of all parameters is enhanced with the increasing signal-to-noise ratio (SNR) and number of snapshots. In addition, they are hardly influenced by the sampling frequency and independent of the initial phase of incident sources. These insights offer guidelines to the system engineer on how to improve parameters’ estimation accuracy.
In this paper, new Cramér-Rao lower bounds (CRB) of the estimates of frequencies, two-dimensional arrival angles and polarization parameters of multiple incident signals are derived for a polarization sensitive array. The incident sources have distinct carrier-frequencies, in contrast to the modeling of all sources to be at the same known carrier-frequency, which has been investigated in the existing research literature on the Cramér-Rao bounds (CRB) for polarization sensitive direction finding. The derived CRBs are compact closed-form expressions and applicable to an arbitrary array geometry. Numerical examples and analysis of some special cases provide insights into the fact that the estimation accuracy of all parameters is enhanced with the increasing signal-to-noise ratio (SNR) and number of snapshots. In addition, they are of influenced by the sampling frequency and independent of the initial phase of incident sources. These insights offer guidelines to the system engineer o n how to improve parameters’ estimation accuracy