基于多接收站的辐射源定位新方法研究

发布时间：2018-05-08 04:29

本文选题：到达时间差 + 到达频率差　；参考：《国防科学技术大学》2014年博士论文

【摘要】：多接收站的辐射源定位是指利用多个接收站接收来自辐射源的信号,从而估计辐射源的位置等信息,简称多站定位。多站定位在电子侦察、环境监测等领域有广泛的应用,如军事中的多站无源定位,民用的无线传感器网络定位等。本文研究的多接收站辐射源定位方法包括了两类定位方法。第一类方法是基于定位观测量的定位方法。定位观测量是指从接收信号中提取的一些含辐射源位置信息的特征量,如到达时间(Time of arrival,TOA)、到达角度(Angle of Arrival,AOA)、到达时间差(Time difference of arrival,TDOA)、到达频率差(Frequency difference of arrival,FDOA)等。基于定位观测量的定位方法首先从接收信号中估计出定位观测量,然后根据定位观测量与辐射源位置之间的观测方程解算出辐射源位置信息。第二类方法是基于信号层次的直接定位方法,这种定位方法不需要从接收信号中估计定位观测量,而是直接从接收信号中估计辐射源的位置。本文第二章至第四章的研究内容都是基于定位观测量的定位方法,其中第二章研究了仅用FDOA的多站定位方法,第三章研究了基于TDOA和FDOA的多站定位与标校方法,第四章研究了基于TOA的多站相对定位方法。第五章是基于信号层次的直接定位方法,研究了基于时延和多普勒的多接收站相干累加直接定位方法。具体研究内容概括如下:第二章研究了仅用FDOA的多接收站辐射源定位方法。首先研究了辐射源存在球面约束时,仅用FDOA的定位方法。本文先分析了球面约束场景下,仅用FDOA定位的克拉美罗下限(Cramer-Rao lower bound,CRLB),然后通过引入拉格朗日乘子将定位问题转换成非线性约束条件下的线性最小二乘估计,最后采用高斯-牛顿方法迭代求解辐射源位置。其次,本章研究了没有球面约束场景下,仅用到达频率差定位时,接收站位置误差对定位精度的影响。本文首先分析了接收站位置误差有无情况下CRLB的不同,通过理论推导分析出接收站位置误差对定位精度影响程度。最后提出一种考虑接收站位置误差因素的定位方法,该方法在FDOA测量噪声和接收站位置误差较小情况下能够达到定位的CRLB。第三章研究了基于TDOA和FDOA的多站定位与标校方法。本章分别研究了三种定位场景下的标校方法。首先是单个辐射源场景下采用单个标校站的定位与标校方法。其次是多辐射源场景下,研究了部分辐射源具有先验信息(等同于含有位置误差的标校站)时的定位与标校方法。最后研究了采用自标校站的多站定位与标校方法。在各自的定位场景下,本文首先分析和推导辐射源定位的CRLB,从而得到采用标校站、辐射源先验位置信息或者自标校站时,辐射源定位精度提高的理论程度。然后提出各自场景下的定位与标校方法,第一步由标校站、含先验位置信息的辐射源或者自标校站的信号到达各个接收站的TDOA和FDOA来估计接收站的位置误差,从而修正接收站的位置和速度。再利用来自辐射源信号的TDOA和FDOA估计辐射源位置和速度。理论分析和实验仿真均表明当TDOA/FDOA测量噪声和接收站位置误差较小情况下,各种定位场景下的定位与标校方法,辐射源定位性能均能达到各自的CRLB。第四章研究了基于TOA的相对定位方法。该定位场景下辐射源和接收站的位置都是未知,因此仅用TOA观测量无法估计出辐射源和接收站的位置,但是各个辐射源和接收站之间的相对位置是可以估计。本文首先推导相对定位场景下的CRLB,通过理论分析得到相对定位问题的衡量标准。其次采用双线性方法将TOA观测量和辐射源位置之间的方程转换为线性关系。通过对TOA观测量矩阵的奇异值分解,将求解辐射源位置的问题转换为求解奇异值分解中的混合矩阵问题。从而将高维数未知量求解问题转换为混合矩阵中九个未知元素的求解问题,降低了问题的复杂度。之后论文建立以混合矩阵元素为未知量的观测方程,通过最小二乘估计方法,求出混合矩阵中的各个元素,进而得到辐射源和接收站的相对位置。然后论文分析了以混合矩阵元素为未知量的观测方程的误差特性,推导出加权矩阵,通过加权最小二乘估计,最终使得定位精度得到进一步提高。最后,论文以双线性方法的估计结果作为初始值,提出一种高斯-牛顿方法迭代求解辐射源的相对位置,从而使得定位精度达到CRLB。第五章研究了基于时延和多普勒的相干累加直接定位方法。论文分别研究了在合作式辐射源定位场景和非合作式辐射源定位场景下,接收站接收多段数据时的定位方法。通过研究不同段接收数据之间相位的差异,以及该相位差值和辐射源位置之间的关系,提出一种基于相干累加的直接定位方法。为了便于对比,论文同时推导了基于非相干累加的直接定位方法。此外,论文推导出这两种定位方法的CRLB。通过CRLB对比表明相干累加直接定位方法的理论定位性能要优于非相干累加直接定位方法。蒙特卡洛仿真表明,在高信噪比下相干累加直接定位方法性能明显优于非相干直接定位方法,而且能够达到相干累加直接定位方法的CRLB。在低信噪比下,两种方法定位性能相似。
[Abstract]:The location of the radiant source of a multi receiving station is to use multiple receiving stations to receive signals from the source, and to estimate the location of the radiation source, for short, multi station positioning. The multi station location is widely used in the fields of electronic reconnaissance and environmental monitoring, such as the multi station passive localization in the military and the location of the civil wireless sensor network. The method of locating the radiant source of the multiple receiving station includes two kinds of location methods. The first class method is based on the positioning method based on the location measurement. The location measurement refers to some characteristics of the source location information extracted from the received signal, such as the arrival time (Time of arrival, TOA), the arrival angle (Angle of Arrival, AOA), and the arrival time. The difference (Time difference of arrival, TDOA), the arrival frequency difference (Frequency difference of arrival, FDOA), etc.. Based on the positioning method, the positioning method first estimates the location measurement from the received signal, and then calculates the location information of the radiation source according to the observation equation between the location measurement and the location of the radiation source. The second method is the base. The method of direct positioning on the signal level, this method does not need to estimate the location measurement from the received signal, but directly estimates the location of the radiation source from the received signal. The second chapter to the fourth chapter of this paper is based on the positioning method based on the location measurement, and the second chapter studies the multi station location method using FDOA only. In the third chapter, the multi station location and calibration method based on TDOA and FDOA is studied. The fourth chapter studies the multi station relative positioning method based on TOA. The fifth chapter is the direct location method based on the signal level, and studies the coherent accumulative direct location method based on time delay and Doppler. The specific content is summarized as follows: the second chapter studies In this paper, the location method of the radiant source in the multi receiving station only using FDOA is studied. First, the location method of FDOA is used only when the radiation source has spherical constraint. This paper first analyzes the Cramer-Rao lower bound, CRLB, which is located in the sphere constrained scene, and then converts the positioning problem into nonlinear by introducing the Lagrange multiplier. The linear least square estimation under the constraint conditions and the Gauss Newton method are used to solve the location of the radiant source iteratively. Secondly, this chapter studies the influence of the location error of the receiving station on the positioning accuracy when the arrival frequency difference is located without the spherical constraint. This paper first analyzes the difference of the location error of the receiving station with or without the CRLB. Through theoretical deduction, the influence degree of location error on the location accuracy is analyzed. At last, a positioning method considering the location error of the receiving station is put forward. This method has studied the multi station location based on TDOA and FDOA based on TDOA and FDOA in the third chapter of the location error under the small error of the measurement noise and the position error of the receiving station. In this chapter, the calibration method of three location scenarios is studied. First, the location and calibration method of a single calibration station is used in a single radiation source scene. Secondly, the location and calibration method of a partial radiant source with a priori information (equivalent to a calibration station with position error) is studied. Using the multi station positioning and calibration method of self calibration station, this paper first analyzes and deduces the CRLB of the location of the radiation source under the respective positioning scene, and then obtains the theoretical degree of improving the location accuracy of the radiation source when the calibration station, the prior location information of the radiation source or the self calibration station, and then the positioning and calibration methods under their respective scenes are put forward. The first step is to estimate the position and speed of the receiving station by the radiation source or the signal of the prior position information to the TDOA and FDOA of the receiving stations to estimate the position and speed of the receiving station. Then the location and speed of the radiation source are estimated by the TDOA and FDOA from the source signal. When the TDOA/FDOA measurement noise and the location error of the receiving station are small, the location and calibration method of various positioning scenes, the location performance of the radiant source can reach the respective CRLB. fourth chapter, and the relative positioning method based on TOA is studied. The location of the radiation source and the receiving station is unknown under this positioning scene, so only TOA is used to measure the location. The location of the radiation source and receiving station is estimated by the method, but the relative position between the sources and the receiving stations can be estimated. Firstly, the CRLB in the relative positioning scene is derived, and the relative positioning problem is obtained by theoretical analysis. Secondly, the bilinear method is used to transform the equation between the TOA observation and the location of the source. By the singular value decomposition of the TOA observation matrix, the problem of solving the location of the radiation source is converted into a mixed matrix problem in singular value decomposition. The problem of solving the problem of the unknown quantity of high dimension is converted to the solution of 9 unknown elements in the mixed matrix, and the complexity of the problem is reduced. After that, the paper is established to mix the problem. The matrix element is the observation equation of the unknown quantity. Through the least squares estimation method, the elements in the mixed matrix are obtained, and the relative position of the radiation source and the receiving station is obtained. Then the paper analyzes the error characteristics of the observation equation with the unknown quantity of the mixed matrix element, and derives the weighted matrix, and the weighted least squares estimate is derived. Finally, the positioning accuracy is further improved. Finally, the paper takes the estimation result of bilinear method as the initial value, and proposes a Gauss Newton method to solve the relative position of the radiation source iteratively, thus making the positioning accuracy up to CRLB. fifth chapter to study the coherent accumulative direct location method based on time delay and Doppler. The location method of the receiving station receiving multiple segments of data in a cooperative radiant location scene and a non cooperative radiation source positioning scene is studied. By studying the phase difference between the received data and the relationship between the phase difference and the location of the source, a direct location method based on coherent accumulation is proposed. In contrast, the paper derives the direct location method based on the incoherent accumulation. In addition, the paper derives the CRLB. of these two positioning methods through the CRLB contrast showing that the theoretical positioning performance of the coherent cumulative direct location method is better than the non coherent cumulative direct location method. Mont Carlo simulation shows that the coherent accumulation of coherent accumulation is directly under high signal to noise ratio. The performance of the joint positioning method is obviously superior to the non coherent direct location method, and the two methods can achieve the similar location performance of the two methods under the low signal to noise ratio of the coherent cumulative direct location method.

【学位授予单位】：国防科学技术大学
【学位级别】：博士
【学位授予年份】：2014
【分类号】：TN95

【参考文献】