An iterated adaptive large neighborhood search algorithm for the large-scale communication satellite range scheduling problem

Abstract

The communication satellite range scheduling problem (CSRSP) is indispensable for the regular operation of the low earth orbit internet constellation, which involves scheduling tracking telemetry and command (TT&C) tasks within their executable arcs to maximize the profit from these scheduled tasks. Different from traditional SRSP, the inter-satellite links are taken into account in CSRSP to facilitate the rapid completion of TT&C tasks. Moreover, the increasing number of satellites and the emergence of associated diverse types of TT&C tasks further escalate the complexity of this problem. Thus, we propose an iterated adaptive large neighborhood search algorithm (IALNS) to solve the CSRSP quickly and straightforwardly. In this algorithm, ALNS is employed to refine heuristic initial solutions. Frequent pattern mining, a popular data mining method, is used to guide the algorithmic search process as iterative mechanisms: on the one hand, the inferior structures in low-quality solutions are mined to significantly assist the ALNS removal process. On the other hand, the superior structures in high-quality solutions are identified to guide the construction of new solutions. Experimental tests with different task scales demonstrate that IALNS effectively deals with the CSRSP, outperforming three state-of-the-art algorithms.