Coupled Communication-navigation Task Scheduling for Lunar Constellations under Load Balance

To address the device-level resource coupling arising from simultaneously running communication and navigation services on shared hardware in lunar satellite constellations, a coupled task scheduling framework based on mixed integer linear programming (MILP) with load-balance-driven optimization was proposed. In this study, heterogeneous resource consumption was transformed into a unified scalar load metric through weighted aggregation, and instantaneous mutual exclusion constraints were converted into cumulative capacity constraints via a two-hour temporal segmentation. A multi-objective optimization model incorporating service quality, task completion rate, navigation efficiency, and load variance was formulated. Load variance minimization served as the core mechanism, implicitly preserving the multi-satellite windows required for navigation through spatial and temporal task redistribution. The quadratic variance objective was exactly linearized into a standard MILP form using L1-norm auxiliary variable decomposition. Simulation experiments were conducted on an eight-satellite elliptical frozen orbit constellation across task scales. The results demonstrated that coupled planning improved the overall task completion rate by an average of 12.3%. The load balance coefficient of variation was consistently superior under coupled planning across all scales.