A Software-Defined Networking Architecture for Aerial Network Optimization

The dynamic nature of aerial networks often results in poor network performance from an end-user perspective due to frequent and extended link outages. With growing expectations of network availability and performance, there is an unmet need to utilize all the available information in the network environment for optimizing network decisions. We present the preliminary design of a Software Defined Networking (SDN)-based prototype system that uses the knowledge about aerial node orbits to predict future network events, especially link outages. By proactively mitigating the effects of network events through SDN-based optimizations, our prototype system is expected to increase the availability and performance of aerial networks. We expect that our generalized approach of using SDNs for leveraging operationally available information for network optimization, is applicable in many different contexts in addition to aerial networking.