节点文献
卫星导航接收机数字波束形成关键技术研究
Study on Key Techniques of Digital Beamforming in Satellite Navigation Receivers
【作者】 李敏;
【导师】 王飞雪;
【作者基本信息】 国防科学技术大学 , 信息与通信工程, 2011, 博士
【摘要】 干扰和多径是卫星导航接收机面临的两大难题。将数字波束形成技术应用于卫星导航接收机,可根据信号、干扰以及多径的不同空间特征,灵活调整阵列加权以形成指向卫星的主波束以及对准干扰、多径的“零陷”,从而达到增强信号、抑制干扰及多径的目的。卫星导航接收机数字波束形成技术的工程实现涉及多个方面的问题,本文以提高卫星信号的接收质量为根本目标,从以下几个方面展开了研究:(1)最优阵列加权准则是数字波束形成技术的核心部分,它是确定阵列加权的理论依据,并决定了天线阵接收机可达到的最佳性能。“最优”是针对某一个优化目标而言的,优化目标不同,最优加权准则的设计结果也不同。卫星导航领域常用的最优准则包括功率倒置准则、最大信干噪比准则、最小均方误差准则等。上述准则对天线阵多径抑制能力的考虑较少,针对卫星导航系统监测站(参考站)典型多径环境,论文提出了一种具有较强多径抑制能力的准则——约束下上比的最小方差(DCMV)准则,仿真结果表明,DCMV准则可在地面反射多径方向形成-20dB或更深的零陷。另外,针对高精度民用接收机对连续波干扰的易感性,论文提出了一种波束/零陷交替控制(SBN)准则,该准则根据不同频率的连续波干扰对接收机的影响程度,交替使用波束控制或零陷控制准则,最大限度地提高了连续波干扰下民用信号的接收质量,实验结果表明,SBN准则下的阵列输出载噪比相比单独使用波束控制准则或零陷控制准则提高了2~5 dB。(2)先验信息误差是影响天线阵波束指向以及抗干扰抗多径性能的重要因素。多种阵列权值优化准则均需要接收机先验信息的辅助,如波束控制准则、最大信干噪比准则、DCMV准则、SBN准则等。针对卫星导航系统,天线阵接收机的先验信息包括星历数据、接收机位置、天线阵平台姿态以及各个阵元的相对位置。传统文献多关注各种类型先验信息误差或其组合的校正方法,缺乏针对具体应用背景的误差分析。勿庸置疑,先验信息误差种类越多,校正过程越复杂,所需要的校正环境也越苛刻,给数字波束形成技术的工程实现带来了较大的困难。论文首次针对卫星导航系统天线阵接收机的各项先验信息误差进行了全面定量分析,其中星历数据、接收机位置以及阵列平台姿态造成的信号到达角误差通常小于1o,而阵元位置安装误差通常不到卫星导航信号载波波长的2%,上述误差对天线阵波束指向以及各项阵列性能的影响均很小,可以不予校正,为数字波束形成技术的工程设计提供了理论依据。此外,还研究了利用导航信号测量阵列平台姿态的方法,给出了有利于提高姿态估计精度的阵型和基线选择建议。(3)阵列通道失配是数字波束形成技术在工程实现中必须要考虑的问题。论文根据天线和射频前端的实测数据,提出了一种符合实际测量结果特点的通道失配仿真模型,并定义了通道间幅度失配、相位失配以及平均群时延差三个参数,定量分析了通道失配对阵列性能的影响。针对卫星导航天线阵接收机,传统的通道校正技术存在明显缺陷,例如针对射频前端的阵列通道失配校正技术无法解决天线之间频率特性的不一致;而在微波暗室对天线阵通道特性进行测量的方法虽然可以校正包括天线在内的通道失配问题,但校正误差可能随器件老化存在漂移,而且需要严格控制校正过程中可能产生的系统误差。在数字信号处理能力大大提高的条件下,利用导航信号进行在线校正是一种理想的校正方法。校正阵列通道失配的关键在于通道特性的估计,基于空时加权结构的通道特性在线估计方法需要多次改变空时加权系数,在不同的空时加权系数下对接收的导航信号进行相关捕获,计算量大且运算时间较长,本文在其基础上进行了改进,只需要一组空时加权系数,通过在相关峰附近进行多次取值代替测量多组相关峰,进而完成对通道特性的估计,大大减小了校正过程的复杂度,节省了运算时间。(4)随着GPS现代化以及Galileo、北斗二号等卫星导航系统的建设,导航用户将会收到更多卫星信号,而接收机的跟踪通道数是有限的,因此选星成为接收机处理的一个重要环节。波束形成下的选星除了需要考虑卫星几何分布对定位解的影响,还需要考虑干扰条件下对卫星信号进行波束增强的可行性。基于上述应用需求,论文提出了一种波束形成下的最优选星方法,以阵列输出信干噪比加权的几何精度因子为选星依据,并针对天线阵接收机的两种结构——单加权网络和多加权网络,提出了最优选星方法的实现流程。基于七阵元天线阵接收机的仿真结果表明:一个宽带干扰时,常规选星方法和本文提出的最优选星方法的定位成功率分别为98.8%和100%;三个宽带干扰时,两种方法的定位成功率分别为91.5%和99.9%,此外,本文提出的方法在100%和99.9%的场景中的平均定位误差与常规选星方法在98.8%和91.5%的场景中的平均定位误差不相上下。可见,波束形成下的最优选星方法既提高了定位成功率,又保证了定位精度。本文的研究成果可用于卫星导航系统监测站接收机、卫星导航增强系统参考站接收机以及高精度航空用户接收机等领域。
【Abstract】 Interference and multipath are two difficult problems confronted by satellite navigation receivers. While digital beamforming (DBF) technology is implemented in satellite navigation receivers, according to different spatial features of signal, interference and multipath, the array weights are adjusted to form main beams towards satellites and nulls towards interference plus multipath. So the purpose of enhancing signals and suppressing interference plus multipath are reached. The engineering implementation of DBF technology in satellite navigation receivers relates to many aspects. This dissertation studied the following aspects, aiming at improving the quality of the received satellite signals.(1) Optimal weighting criterion is the key part of DBF technology. It constructs the theoretical foundation of computing the array weights, and decides the optimal performance of an array receiver.“Optimal”refers to certain optimization target. While the optimization target differs, the designing result of the optimal array criterion is also different. Optimal criteria mostly used in satellite navigation systems include Power Inversion, Maximum Signal-to-Interference-plus-Noise Ratio, Minimum Mean Square Error, etc. Those criteria considered few aspects of suppressing multipath. According to the typical multipath environment of monitoring stations (reference stations) in satellite navigation systems, this dissertation proposed an optimal criterion to mitigate multipath, which is named Down-up-ratio Constrained Minimum Variance (DCMV) criterion. Simulation results showed that DCMV criterion could form a -20dB or even deeper null in the direction of multipath reflected from the ground. Since high precision civil receivers are vulnerable to continuous wave interference, this dissertation also proposed a criterion named Switchable Beam-steering/ Null-steering (SBN) criterion. It switches Beam Steering criterion to Null Steering criterion or inversely according to the impact of the continuous wave interference under different frequency. This criterion can extremely improve the quality of the received civil signals interfered by continuous wave. Experiment results showed that the array output carrier-to-noise-ratio under SBN criterion got 2~5 dB improvement compared to the one under Beam Steering criterion or Null Steering criterion separately.(2) The error of prior information is an important factor which affects the beam direction of the antenna array and the performance of suppressing interference and multipath. Many optimal criteria need the aid of prior information, such as Beam Steering, Maximum Signal-to-Interference-plus-Noise Ratio, DCMV, SBN, etc. In satellite navigation systems, the prior information of antenna array receivers includes the ephemeris data, the location of the receiver, and the platform attitude of the antenna array plus the relative positions of the array elements. Traditional literature usually focused on the calibration of one kind or the combined errors of the prior information, but neglected the error range of the prior information based on the application. Inevitably, the more the types of prior information are, the more complicated the calibrating procedure is, and the stricter the calibrating environment is needed, which may cause many difficulties in the implementation of DBF technology. This dissertation analyzed the error range of each type of the prior information in satellite navigation antenna array receivers for the first time. The estimation error of the signal arriving angle caused by errors from the ephemeris data, the location of the receiver and the platform attitude of the antenna array is less than 1 degree; Position errors of the array elements are usually less than 2% of the signal carrier wave length. The above errors have little impact on the beam direction and other array performance, and need no calibration. This conclusion provides the theoretical reference to the engineering implementation of DBF technology. Also, the platform attitude calibration with navigation signals was studied, and the suggestion about array configuration and baseline selection were given.(3) Array channel mismatch must be considered in the engineering implementation of DBF technology. This dissertation proposed a channel mismatch model according to the measured data of antennas and RF front ends, then defined three parameters: the amplitude mismatch, the phase mismatch, the difference between averaged group delays. And the affect of channel mismatch on array performance were analyzed in quantity. For antenna array satellite navigation receivers, traditional channel mismatch calibration has obvious disadvantages, for example, it is not able to solve the mismatch between antennas; the method of calibrating antenna array channel mismatch in anechoic chambers is able to calibrate mismatches between antennas, but the calibrating error may float when the hardware gets old, also the system error in the calibrating procedure should be controlled strictly. Nowadays, digital signal processing ability has been improved extremely, online calibrating with satellite signal is an ideal method. The key part of calibrating channel mismatch is to estimate the channel characteristic. Based on the space-time structure, the channel characteristic estimation has to change the space-time weights and capture the correlation peaks under each group of space-time weights; which costs large computation and time. This dissertation improves the online calibrating method, which only needs one group of space-time weights. The proposed method measures several values near the correlation peak, instead of measuring several correlation peaks, and then estimates the channel characteristic. The proposed method reduced the complexity of calibration, and also saved the computation time.(4) Along with the construction of the modern GPS, Galileo and Compass II, GNSS users will receive more satellite signals, but the amount of tracking channels of the receiver is limited. Therefore, satellite selection becomes an important step in receiver’s processing. Satellite selection under DBF needs to consider not only the impact of satellite geometry, but also the feasibility of enhancing satellite signals with array beams. According to the above requirement, this dissertation proposed the optimal satellite selecting method for beamforming. Satellites are selected according to the geometry dilution of precision weighted by the array output signal to interference-plus-noise ratio. Based on the single-weight structure and the multi-weights structure of the antenna array receivers, the flow charts of the optimal satellite selecting procedure are suggested. Simulations are conducted to compare the performance of the conventional method and the proposed method. It shows that for typical seven-element antenna array receivers, the positioning success ratio is 98.8% and 100% respectively when there is one broadband interference; and the positioning success ratio is 91.5% and 99.9% when there are three broadband interference. Besides, the average positioning error of the proposed method in the 100% and 99.9% scenarios is almost the same with that of the conventional method in the 98.8% and 91.5% scenarios. It can be seen that the optimal satellite selecting method for beamforming improves the positioning success ratio, and at the mean time guarantees the positioning accuracy.The technique studied in this dissertation can be applied in the monitoring receivers of satellite navigation systems, the reference receivers of satellite navigation augmentation systems and the receivers for high-precision aerospace users.
【Key words】 Satellite Navigation Receiver; Antenna Array; Digital Beamforming; Interference Mitigation; Multipath Mitigation; Pattern; Array Gain; Down Up Ratio; Prior Information; Channel Mismatch; Positioning Accuracy; Satellite Selection; Geometry Dilution of Precision;