Software-defined networking (SDN) logically separates the control and data planes, thus opening the way to more flexible configurations and management of low-Earth orbit (LEO) satellite networks. Since one or, more generally, multiple distributed controllers are needed, a significant challenge in SDN is the controller placement problem (CPP). Due to characteristics such as the dynamic network topology, limited bandwidth and traffic variations, the CPP is quite complex in SDN-based satellite networks. In this paper, we propose solving the CPP by means of a static placement with dynamic assignment (SPDA) method for LEO satellite networks. The SPDA method has two parts: the first is to incorporate SDN controllers into some fixed satellites by formulating a mixed integer programming model; the second is to dynamically assign switches to existing controllers according to the switch-controller latency and the traffic load of controllers. The SPDA method takes the topological dynamics into account by effectively dividing time snapshots, and it has a lower bandwidth consumption compared with methods involving controller migrations. Real satellite constellations are used to evaluate the performance of our controller placement solution. The results show that SPDA outperforms existing methods in terms of reducing the switch-controller latency, and it also has good load balancing performance.
Mega-constellation networks, represented by Starlink and OneWeb, have become a promising solution for the wide-area Internet of Things (IoT). IoT messages collected by satellite can be routed to the ground gateway via multiple intersatellite link (ISL) relays and then access the ground network. Compared to traditional constellations, mega-constellations with massive satellites require more ISL relays that are greatly affected by the number and location of ground gateways. In this paper, we focus on the ISL usage for connecting satellites and gateways and propose a novel method with low computation cost to evaluate the ISL usage metric. Then we formulate a mixedinteger optimization model for the gateway site optimization (GSO) problem with minimizing the overall ISL usage, which is then simplified through model transformation. And an IBD-PSO algorithm is proposed to solve the transformed GSO problem. Based on the Starlink constellation, simulation results have verified the proposed method by comparisons with previous studies. We further investigate how the gateway placement is affected by gateway number and traffic demand pattern. The relations between gateway placement and ISL hop-count of different satellites are also studied.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.