Adaptive GNSS-5G hybrid positioning based on time offset optimization estimation and multi-rate measurement fusion

Abstract

The fusion positioning of the Global Navigation Satellite System and Fifth-Generation Mobile Communication Network is a key direction for breaking through the performance bottleneck of a single system. However, it faces two core challenges: inconsistent spatiotemporal benchmarks and multi-rate measurement fusion. To address these issues, this paper proposes a joint time offset estimation and phased fusion strategy: An adaptive time-varying offset model is established, and an adaptive relative time offset estimation algorithm based on pseudo-measurements is designed. The high-precision time benchmark provided by GNSS is used to realize the indirect estimation of the 5G absolute time offset, solving the problem of offset accumulation in dynamic scenarios. A two-stage filtering framework is proposed, which processes the coordinate conversion error of 5G polar coordinate measurements through a modified unbiased converted measurement Kalman filter, combines with a Kalman filter to estimate the target state, and constructs a time offset pseudo-measurement based on velocity estimation for efficient solutions. A phased multi-rate fusion strategy is designed: At GNSS sampling moments, adaptive weighted fusion is used to correct the accumulated errors of 5G high-frequency data; at non-GNSS moments, 5G high-frequency measurements and motion state equation predictions are used to maintain tracking accuracy for high-dynamic targets. Simulation results show that the proposed algorithm significantly outperforms eight mainstream algorithms such as SPP, EKF, and UKF in positioning accuracy, with a total average error of 1.66 m and a total root mean square error of 2.02 m. Moreover, the error distribution is more concentrated and stability is stronger, which can effectively adapt to the needs of high-dynamic scenarios and provide reliable solutions for GNSS-5G hybrid positioning.