In this paper we study blind source localization problem based on the joint received signal strength difference (RSSD) and angle of arrival (AOA) measurements with unknown transmit power of source. Since RSSD and AOA measurements are uncorrelated, combining two methods leads to a better performance for source localization. This paper focus on the pseudo linear estimator (PLE) with a closed-form and low complexity solution. One of the main limitations in this estimator is the bias created from the correlation between system matrix and error vector, which is not vanished by increasing the number of measurements. To overcome this problem first, we present a bias compensated PLE using the closed instrumental variable (IV). Then, for improving the localization performance a weighting IV estimator (WIV) is presented. Finally, for achieving the Cramer–Rao lower bound (CRLB) an improved WIV (IWIV) estimator is used based on the known relation between the estimated parameters of WIV estimator. The proposed IWIV estimator is proved to be asymptotically efficient (i.e., obtaining zero bias and the Cramer–Rao lower bound). Numerical simulations also verify the theoretical development and show source localization using hybrid information RSSD/AOA has a superior performance than RSSD and AOA solely.

在本文中，我们基于未知源发射功率的联合接收信号强度差（RSSD）和到达角（AOA）测量，研究了盲源定位问题。由于RSSD和AOA的测量是不相关的，因此将两种方法结合使用可提高源定位的性能。本文重点介绍具有封闭形式和低复杂度解决方案的伪线性估计器（PLE）。该估计器的主要限制之一是由系统矩阵和误差向量之间的相关性产生的偏差，该偏差不会因增加测量次数而消失。为了首先克服这个问题，我们提出了一种使用闭式工具变量（IV）的偏置补偿PLE。然后，为了提高定位性能，提出了加权IV估计器（WIV）。最后，为了实现Cramer-Rao下界（CRLB），基于WIV估计器的估计参数之间的已知关系，使用了改进的WIV（IWIV）估计器。所提出的IWIV估计器被证明是渐近有效的（即，获得零偏差和Cramer-Rao下界）。数值模拟也验证了理论发展，并显示了使用混合信息RSSD / AOA进行源定位比单独使用RSSD和AOA具有更好的性能。