Reliable real-time triple-frequency cycle slip detection and recovery with adaptive detection thresholds

Abstract

Fast and reliable cycle slip detection is a prerequisite for successive ambiguity resolution and precise positioning. For dual-frequency signals, sometimes small cycle slips or particular cycle slips cannot be detected. Benefiting from the triple-frequency signals of global navigation satellite systems, more useful linear combinations can be selected to reduce or avoid the cycle slips otherwise difficult to detect. We propose an improved algorithm with three independent geometry-free combinations to detect and recover potential cycle slips. To improve the reliability of cycle slip detection, adaptive thresholds determined using the posterior total noise of the approximately normal-distributed time-differenced combinations are used, rather than fixed thresholds by a priori information such as the measurement noise and ionosphere residual. After the detection procedure, four different types of detection results are obtained, and the corresponding recovery strategies are applied. In this way, we can reliably detect and recover potential integer cycle slips. To demonstrate the performance of the proposed algorithm, numerical analysis is carried out using observation data of the BeiDou Navigation Satellite System with real cycle slips. Different types of cycle slips are also discussed. The results show that the proposed algorithm can reliably avoid false-acceptance or false-rejection caused by the fixed thresholds, and correctly detect and recover different kinds of integer cycle slips for the experimental data with a 1 s sampling rate.