A binary discrete particle swarm optimization satellite selection algorithm with a queen informant for Multi-GNSS continuous positioning. Issue 9 (1st November 2021)

• Record Type: Journal Article
• Authors: Zhao, Dongqing; Cai, Congcong; Li, Linyang
• Is Part Of: Advances in space research. Volume 68:Issue 9(2021)
• Page Start: 3521
• Page End: 3530
• Publication Date: 2021-11-01
• Subjects: Multi-GNSS -- Continuous positioning -- Satellite selection -- EPSO with a queen informant -- Geometric dilution of precision; Space sciences -- Periodicals; Astronautics -- Periodicals; Geophysics -- Periodicals; 500.505
• DOI: 10.1016/j.asr.2021.08.013
• Languages: English
• ISSNs: 0273-1177

Abstract

Abstract: Currently, there are more than 130 available navigation satellites, a much larger scale low earth orbit (LEO) satellites will be deployed in the near future. However, limited by computational performance, capacity and receiver channels, a subset of all visible satellites is suggested to be selected with a with better geometric dilution of precision (GDOP). Since a reduced set of used satellite will have little effect on positioning accuracy, in contrast, higher real-time performance will be obtained. In view of the problems of the particle swarm optimization (PSO) algorithm for satellites selection, we adopt an easy binary particle swarm optimization with a queen informant (EPSOq) algorithm. This avoids the concept of "speed" in the discrete PSO algorithm and directly calculates the probability of position value when updating particles, with a queen particle used to accelerate convergence. Further, considering the continuity of positioning, we use a sequence to determine whether satellite re-selection is necessary. If the visible satellite has not changed greatly, the results of the last satellite selection is used instead of selecting new satellites. The observation data collected in the East China Sea is utilized in the experiment. Compared with the traversal method which requires 51.215 s, the average calculation time of EPSOq-C is only 0.0025 s with a population size of 70, and the calculation speed is an order of magnitude faster than the PSO algorithm. Furthermore, the distribution of GDOP bias derived from EPSOq-C is more concentrated than the PSO algorithm.