The time-hopping direct sequence spread spectrum (TH-DSSS) signal has been widely used in Pseudolites Positioning Systems to overcome the near-far problem. To capture the TH-DSSS signal, an additional parameter representing the time-hopping (TH) rules should be estimated in addition to the PRN code phase and carrier Doppler. However, the techniques of estimating a TH parameter in existing TH-DSSS signal acquisition methods have significant issues in poor signal quality environments. Here, we propose a robust and general TH-DSSS signal acquisition method to reduce the impact of signal degradation. In this method, we first capture every short pulse to obtain the code phase and carrier Doppler. After sufficient successful pulse acquisitions, we model the process of TH parameter acquisition as a dynamic Bayesian network. The so-called state confidence that describes the probability of each candidate TH parameter is then introduced to infer the real TH parameter. Finally, this method has been seen, both theoretically and experimentally, to be both general and effective to compensate for harsh signal environments. Simulation results show that compared with baseline algorithms, this method provides a significant improvement in detection probability and considerable reduction in acquisition time.

跳时直接序列扩频（TH-DSSS）信号已广泛用于伪卫星定位系统中，以解决近距离问题。为了捕获TH-DSSS信号，除了PRN码相位和载波多普勒信号外，还应该估算一个代表时间跳变（TH）规则的附加参数。但是，在现有的TH-DSSS信号采集方法中，估计TH参数的技术在信号质量较差的环境中存在重大问题。在此，我们提出了一种鲁棒且通用的TH-DSSS信号采集方法，以减少信号降级的影响。在这种方法中，我们首先捕获每个短脉冲以获得码相位和载波多普勒。成功进行足够的脉冲采集后，我们将TH参数采集过程建模为动态贝叶斯网络。然后引入描述每个候选TH参数概率的所谓状态置信度以推断真实TH参数。最后，从理论上和实验上都可以看出这种方法对于补偿恶劣的信号环境是通用且有效的。仿真结果表明，与基线算法相比，该方法显着提高了检测概率，并大大减少了采集时间。