In ground station scheduling problem the aim is to compute an optimal planning of communications between Spacecrafts (SCs) and operations teams of Ground Stations (GSs). While such allocation of tasks to ground stations traditionally is mostly done by human intervention, modern scheduling systems look at optimization and automation features. Such features, on the one hand, would increase the efficiency and productivity of the mission planning systems by handling a larger number of missions, achieve a higher usage of the infrastructure (grand stations' antennae) and, on the other, would avoid error-prone human allocation and reduce human labour costs. Designing such modern, automated scheduling/planning systems is however challenging due to the highly constraint and complex nature of the problem seeking to optimize along various objectives or system parameters. In this paper we present a study on the performance of several meta-heuristics methods for solving ground station scheduling problem. Local search methods (Hill Climbing, Simulated Annealing and Tabu Search) and population-based methods (GA, Steady State GA and Struggle GA) have been considered for the study. The performance of these resolution methods was measured by a set of instances of varying size and complexity generated by STK toolkit. The study revealed the strengths and weaknesses of the considered methods while solving different size instances and considering several objective functions, namely, windows fitness, clashes fitness, time requirement fitness, and resource usage fitness.

Peer Reviewed