Distributed GNSS Collaborative Localization: Theoretical Performance Analysis and Simulation Verification

In global navigation satellite system (GNSS) collaborative positioning, each user shares its processed data and conducts joint processing. For distributed positioning algorithms, the GNSS users can only get observational data from their neighboring users, thus the performance is relatively worse than the centralized algorithms. However, compared with centralized algorithms, the distributed algorithms are more suitable for practical applications since they do not need a processing center. Thus, the performance analysis of distributed algorithms is essential for the actual applications. In this paper, based on the proposed signal and system model of distributed GNSS collaborative positioning, the Distributed Least Squares (DLS) algorithm and the corresponding Collaborative Dilution Of Precision (CDOP) are introduced in detail. After that, the theoretical analysis and simulation tests are carried out to analyze the performance of distributed algorithms in different scenario, which focus on the positioning availability, computation complexity and the impacts of the number and distribution of collaborative users. Finally, the DLS algorithm is tested in some real scenarios with several real-time GNSS receivers. Theoretical analysis and experimental results both verify that the positioning availability and accuracy can be improved by distributed collaborative algorithms, and the advantages and disadvantages of distributed collaborative algorithms are summarized which can provide some useful information for the design and testing of the actual collaboration applications.