CSAC Characterization and Its Impact on GNSS Clock Augmentation Performance

• Main Authors: Enric Fernández, David Calero, M. Eulàlia Parés
• Format: Article
• Language: English
• Published: MDPI AG 2017-02-01
• Series: Sensors
• Subjects: CSAC; calibration; clock augmentation; GNSS; atomic clock; navigation; steering
• Online Access: http://www.mdpi.com/1424-8220/17/2/370

Chip Scale Atomic Clocks (CSAC) are recently-developed electronic instruments that, when used together with a Global Navigation Satellite Systems (GNSS) receiver, help improve the performance of GNSS navigation solutions in certain conditions (i.e., low satellite visibility). Current GNSS receivers include a Temperature Compensated Cristal Oscillator (TCXO) clock characterized by a short-term stability (τ = 1 s) of 10−9 s that leads to an error of 0.3 m in pseudorange measurements. The CSAC can achieve a short-term stability of 2.5 × 10−12 s, which implies a range error of 0.075 m, making for an 87.5% improvement over TCXO. Replacing the internal TCXO clock of GNSS receivers with a higher frequency stability clock such as a CSAC oscillator improves the navigation solution in terms of low satellite visibility positioning accuracy, solution availability, signal recovery (holdover), multipath and jamming mitigation and spoofing attack detection. However, CSAC suffers from internal systematic instabilities and errors that should be minimized if optimal performance is desired. Hence, for operating CSAC at its best, the deterministic errors from the CSAC need to be properly modelled. Currently, this modelling is done by determining and predicting the clock frequency stability (i.e., clock bias and bias rate) within the positioning estimation process. The research presented in this paper aims to go a step further, analysing the correlation between temperature and clock stability noise and the impact of its proper modelling in the holdover recovery time and in the positioning performance. Moreover, it shows the potential of fine clock coasting modelling. With the proposed model, an improvement in vertical positioning precision of around 50% with only three satellites can be achieved. Moreover, an increase in the navigation solution availability is also observed, a reduction of holdover recovery time from dozens of seconds to only a few can be achieved.