A Back-tracing Partition based On-path Caching Distribution Strategy over integrated LEO Satellite and Terrestrial networks

Abstract

In these years, hybrid low earth orbit (LEO) satellites and terrestrial networks are widely developed for a variety of content distribution services, such as internet of things, satellite sensor networks and tactical distribution networks, due to low deployment cost and broader coverage capability. However, an unreasonable distribution strategy in the integrated LEO satellite and terrestrial user network will incur prolonged content access latency and significant overheads for users, due to obviously time-varying topology and limited on-board resources. In this paper, therefore, we proposed a back-tracing partition directed on-path caching mechanism for the publisher/subscriber distribution over the hybrid LEO constellation and terrestrial network. In the proposed mechanism, a back-tracing algorithm instructed node selection strategy is well-designed as well as a cross-slot graph based Dijkstra routing algorithm. Through avoiding intermittent connectivity as much as possible, the proposed mechanism could efficiently reduce the redundant transmissions of data access for different users, by fetching objective file mainly from a limited group of intermediate caching nodes, instead of directly from the source node. The simulation results verified the proposed method can obviously degrade the overall overheads and access delay of all users compared with the minimum spanning tree (MST) algorithm.