Clock Error Analysis and Compensation for LEO Signal of Opportunity Positioning

Opportunity positioning by low Earth orbit (LEO) satellite signals has been widely researched and developed as a supplement to global navigation satellite system (GNSS). However, poor system clock stability and multiple consecutively received epochs' Doppler positioning method really limit the LEO opportunity positioning accuracy. To address the problems, this article systematically analyzes the best linear unbiased estimation (BLUE) of users' position under clock error and the significance of a priori clock error information in terms of positioning accuracy improvement. Then a novel mutual feedback positioning method based on the close coupling of long short-term memory (LSTM) network and error state Kalman filter (ESKF) is proposed to compensate for the effect caused by clock error. Experiments are conducted through the actual acquisition of Iridium signals by LEO receivers equipped with different performance clocks, and the results show that the LSTM-ESKF positioning method could improve the accuracy by one order of magnitude compared with the traditional method when the clock performance is poor, which verifies the effectiveness of the positioning method. This is important to improve the positioning performance and reduce the cost in clock selection of LEO receivers in real applications.