BOC信号同步算法研究与实现

发布时间：2018-04-24 09:53

本文选题：BOC调制 + SPART算法　；参考：《上海交通大学》2014年博士论文

【摘要】：在现代导航系统中，BOC调制信号的采用使导航系统间实现频谱共享，同时提高定位精度和抗多径干扰的能力成为可能。但是,BOC信号子载波调制所带来的同步模糊性是当今亟待解决的关键热门问题。由于BOC信号自相关函数的多峰值特性，曾经适用于BPSK信号的同步算法已不再适用。本文在深入研究BOC调制信号特性和传统同步方案的基础上，提出了三种解决同步模糊性的算法。提出了SPART(Symmetrical Pulse Ambiguity Removing technique)非模糊性同步算法：算法的原理是构造两个本地BOC-like信号，其中一个是偶对称的，另外一个是奇对称的信号。通过合成这两个本地信号与BOC信号的互相关函数来获得具有非模糊性的相关函数。合成的相关函数仅含有一个正的相关峰值，所有的旁峰都被消除掉了,从而巧妙地消除了旁峰对捕获和跟踪处理的潜在威胁。文中不但给出了SPART算法的实现环路结构，同时从基于Monte Carlo的仿真分析和基于FPGA的硬件平台对SPART算法的有效性进行了验证。对捕获检测概率和跟踪标准差的表达式做了理论上的推导。仿真发现，SPART算法不仅适合正弦BOC信号，也适用于余弦BOC信号，，但针对余弦BOC信号时，SPART算法采用的本地信号的脉宽是针对正弦BOC信号脉宽的一半。SPART算法与传统同步算法对比分析发现，在相同积分时间，码间距和累加次数的情况下，针对低调制系数的BOC信号，SPART算法在捕获和跟踪阶段与传统跟踪环路相比有很小的性能退化。调制系数越高，BOC信号退化越严重。但是不要忘了SPART算法的初衷是消除BOC信号自相关函数中正的侧峰以免捕获到侧峰上和锁定到错误的鉴相器零点上。另外，为了补偿SPART算法的性能退化，我们可以通过增加相干积分时间，增加非相干累加次数和减小码间距的方法来提高算法的性能。提出了适合BOC(n,n)和MBOC信号的GFSA(Gating function Shifting Algorithm)非模糊性算法：算法的原理是构造一个选通脉冲信号去调制本地的PRN码和子载波信号，调制后的本地信号与接收到的BOC进行互相关运算，相关函数向左和向右移位产生的两个相关函数，当选通函数的脉冲宽度在[0.51]个码片范围内时，它们和的能量减去差的能量就可以合成一个仅含有一个正峰值的相关函数。文中给出了GFSA算法的实现环路结构，对GFSA算法的跟踪性能做了理论上的推导和基于Monte Carlo的仿真分析。仿真分析表明，针对BOC(n,n)信号，在相同积分时间，码间距和累加次数的情况下，GFSA算法在跟踪阶段与传统跟踪环路相比依然有性能退化现象。在选通脉宽为1个码片时，最大仅有2.1分贝的性能退化。随着选通脉宽的减小，性能退化越来越不明显，在0.5个码片宽度和码间距为0.2个码片条件下仅有1分贝的性能退化，且退化的绝对值相对于传统跟踪算法非常小。所以，GFSA算法在消除BOC信号同步模糊性的同时利用窄选通函数可以获得很好的跟踪特性。由于滤波效应，采用窄脉宽选通函数是切实可行的。另外，由于设计的跟踪环路不再需要子载波生成电路，所以它简单易行。提出了适合BOC(n,n)和MBOC的BPST(BOC-PRN shifting technique)非模糊性同步算法：BPST算法的基本思想是利用接收到的BOC信号与本地PRN码的互相关函数来合成一个具有非模糊特性的相关函数。BPST算法的接收机采用相同的结构和不同延时的PRN序列就可以接收正弦或余弦BOC信号。对BPST算法的捕获和跟踪性能做了理论上的推导和基于Monte Carlo的仿真分析。理论和仿真分析表明，对中长距离的多径干扰信号有极强的抑制能力。尽管BPST算法，在相同积分时间，码间距和累加次数的情况下，在捕获和跟踪阶段分别有0.5分贝和2.2分贝的性能损失，但它彻底解决了同步的模糊性问题。同其它非模糊性算法相比，性能相差无几。因此，对sinBOC(n,n), cosBOC(n,n)和MBOC信号接收机来讲采用BPST算法不失为一个很好的同步解决方案。本文最大的成果就是提出了BOC信号捕获和跟踪环路及接收机实现结构，以使现在实用的和计划准备使用GNSS接收信号的使用者可以获得更好的定位服务。
[Abstract]:In modern navigation systems, the use of BOC modulation signals enables the sharing of spectrum between the navigation systems, and it is possible to improve the positioning accuracy and the ability to resist multipath interference. However, the synchronization ambiguity caused by the BOC signal subcarrier modulation is the key hot question to be solved urgently. Because of the multiple peak characteristics of the autocorrelation function of the BOC signal The synchronization algorithm used for BPSK signals is no longer applicable. Based on the in-depth study of the characteristics of the BOC modulation signal and the traditional synchronization scheme, three algorithms to solve the synchronization ambiguity are proposed.
A non fuzzy synchronization algorithm for SPART (Symmetrical Pulse Ambiguity Removing technique) is proposed. The principle of the algorithm is to construct two local BOC-like signals, one is even symmetric and the other is a singly symmetric signal. By synthesizing the cross correlation functions of the two local signals and BOC signals, the non fuzzy phase is obtained. The correlation function contains only a positive correlation peak, and all the adjacent peaks are eliminated, which ingeniously eliminates the potential threat to the acquisition and tracking of the side peaks. Not only is the implementation loop structure of the SPART algorithm, but also the simulation analysis based on the Monte Carlo and the SPAR based hardware platform based on the FPGA The validity of the T algorithm is verified. A theoretical deduction is made for the expression of the capture detection probability and the standard deviation of the tracking standard. The simulation shows that the SPART algorithm is not only suitable for sinusoidal BOC signal, but also suitable for the cosine BOC signal, but for the cosine BOC signal, the pulse width of the local signal used by the SPART algorithm is half the pulse width of the sinusoidal BOC signal. Compared with the traditional synchronization algorithm, the.SPART algorithm shows that in the case of the same integration time, the code spacing and the number of accumulations, the SPART algorithm has little performance degradation compared with the traditional tracking loop for the BOC signal with low modulation coefficient. The higher the modulation coefficient, the worse the BOC signal degradation. But don't forget the SPART. The original intention of the algorithm is to eliminate the positive side peaks in the autocorrelation function of the BOC signal in order to avoid capturing the side peaks and locking to the wrong phase detector zero. In addition, in order to compensate for the performance degradation of the SPART algorithm, we can increase the performance of the algorithm by increasing the coherence time, increasing the incoherent accumulating times and reducing the spacing of the small code.
A GFSA (Gating function Shifting Algorithm) non fuzzy algorithm suitable for BOC (n, n) and MBOC signals is proposed. The principle of the algorithm is to construct a gated pulse signal to modulate the local PRN code and subcarrier signal. The modulated local signal is interrelated with the received BOC, and the correlation function is shifted left and right to the left and right. Two correlation functions, when the pulse width of the elected pass function is within the range of [0.51] codes, the energy minus the energy can be reduced to a correlation function that contains only one positive peak. The implementation loop structure of the GFSA algorithm is given in this paper, which can be derived theoretically and based on the Monte Carlo based on the GFSA algorithm. Simulation analysis shows that, for the BOC (n, n) signal, the GFSA algorithm still has the performance degradation phenomenon compared with the traditional tracking loop at the same integration time, the code spacing and the number of accumulative times. The performance degradation is only 2.1 dB when the pulse width is 1 codes, and the performance is reduced with the decrease of the gating pulse width. It is becoming less and more obvious that only 1 decibels are degraded under the condition of 0.5 chip width and code spacing of 0.2 codes, and the absolute value of degradation is very small compared with the traditional tracking algorithm. Therefore, the GFSA algorithm can obtain good tracking characteristics by using narrow pass function while eliminating synchronization ambiguity of BOC signals. The narrow pulse width gating function is feasible. Moreover, since the tracking loop is no longer needed for the subcarrier generation circuit, it is simple and feasible.
A BPST (BOC-PRN shifting technique) non fuzzy synchronization algorithm suitable for BOC (n, n) and MBOC is proposed. The basic idea of BPST algorithm is to use the intercorrelation function of the received BOC signal and the local PRN code to synthesize a receiver with a non fuzzy correlation function.BPST algorithm using the same structure and different delay sequence. A sine or cosine BOC signal can be received. The acquisition and tracking performance of the BPST algorithm is theoretically derived and the simulation analysis based on Monte Carlo is made. The theoretical and simulation analysis shows that the medium and long distance multipath interference signals have a very strong suppression ability. Although the BPST algorithm is at the same integration time, the distance between the code and the number of accumulations In the capture and tracking phase, the performance loss of 0.5 dB and 2.2 DB respectively, but it completely solves the problem of synchronization fuzziness. Compared with other non fuzzy algorithms, the performance is very different. Therefore, it is a good synchronization solution for sinBOC (n, n), cosBOC (n, n) and MBOC signal receivers to adopt BPST algorithm.
The greatest achievement of this article is the proposed BOC signal capture and tracking loop and the receiver implementation structure, so that the users who are now practical and plan to use the GNSS receiving signal can get better positioning service.

【学位授予单位】：上海交通大学
【学位级别】：博士
【学位授予年份】：2014
【分类号】：TN911.3

【参考文献】