Scholarlink

<p>Global Navigation Satellite System (GNSS) satellites are capable of ranging and communicating with one another through Inter-Satellite Links (ISLs), thereby enhancing system autonomy. ISL assignment can be formulated as a complex Constrained Multi-objective Optimization Problem (CMOP). It is critical to GNSS performance but is heavily constrained by compact satellite platforms and point-to-point link types. Prior research has predominantly focused on meeting communication and ranging requirements, often neglecting the load-balancing requirement within the system. To address these challenges, we propose the Constrained ISL-dependent NSGA-II (C-ISL-NSGA-II) method. This algorithm incorporates a novel Constrained Single-Link Rebuilding (CSLR) mutation operator and simultaneously optimizes ranging, communication, and load-balancing performance. Using the BeiDou constellation as a case study, we demonstrate that the CSLR mutation operator outperforms traditional Exchange Mutation (EM) and Displacement Mutation (DM) operators. It achieves faster convergence and superior solution quality across all three objectives. Additionally, we analyze the effects of crossover and mutation rates, finding that a configuration with a high mutation rate and a low crossover rate yields the best Pareto-set approximation for the BeiDou-3 constellation. An empirical&#xa0;parametric study indicates that employing 900 generations with a population of 100 individuals achieves a favorable trade-off among convergence speed, diversity, and computational cost. Compared with randomly generated initial solutions, the mean values of WAPDOP, SDDTD, and WAVWNL in the obtained Pareto-optimal solutions are reduced by 63%, 72%, and 78%, respectively. Finally, by expanding the problem to incorporate a third load-balancing objective, we observe a significant reduction in total link loads and inter-satellite load disparities, albeit at the expense of a slight decrease in ranging accuracy.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Multi-objective optimization of inter-satellite links assignment for GNSS based on constrained non-dominated sorting genetic algorithm

  • 期刊论文

  • 发表日期：

    2026年7月16日

查看全文
  • Xiaoying Gong,
  • Mengyao Chai,
  • Shixiang Cai,
  • Shilong Zhang,
  • Dingfa Huang

摘要

Global Navigation Satellite System (GNSS) satellites are capable of ranging and communicating with one another through Inter-Satellite Links (ISLs), thereby enhancing system autonomy. ISL assignment can be formulated as a complex Constrained Multi-objective Optimization Problem (CMOP). It is critical to GNSS performance but is heavily constrained by compact satellite platforms and point-to-point link types. Prior research has predominantly focused on meeting communication and ranging requirements, often neglecting the load-balancing requirement within the system. To address these challenges, we propose the Constrained ISL-dependent NSGA-II (C-ISL-NSGA-II) method. This algorithm incorporates a novel Constrained Single-Link Rebuilding (CSLR) mutation operator and simultaneously optimizes ranging, communication, and load-balancing performance. Using the BeiDou constellation as a case study, we demonstrate that the CSLR mutation operator outperforms traditional Exchange Mutation (EM) and Displacement Mutation (DM) operators. It achieves faster convergence and superior solution quality across all three objectives. Additionally, we analyze the effects of crossover and mutation rates, finding that a configuration with a high mutation rate and a low crossover rate yields the best Pareto-set approximation for the BeiDou-3 constellation. An empirical parametric study indicates that employing 900 generations with a population of 100 individuals achieves a favorable trade-off among convergence speed, diversity, and computational cost. Compared with randomly generated initial solutions, the mean values of WAPDOP, SDDTD, and WAVWNL in the obtained Pareto-optimal solutions are reduced by 63%, 72%, and 78%, respectively. Finally, by expanding the problem to incorporate a third load-balancing objective, we observe a significant reduction in total link loads and inter-satellite load disparities, albeit at the expense of a slight decrease in ranging accuracy.