An SDN Based Framework for Load Balancing and Flight Control in UAV Networks

Abstract

Unmanned Aerial Vehicles (UAVs) are gaining tremendous attention due to their flying nature. To complete the task efficiently, multi UAV systems are a good choice as compared to a single UAV system. However, multi-UAV systems introduce issues such as high dynamics, limited battery, and frequent changes in topology. Software control is required to solve these issues. Thus, Software Defined Networking (SDN) is an excellent candidate to separate control logic from forwarding elements and provide high-level programming abstractions. However, due to architectural constraints, applying SDN introduces some new challenges, including uneven load on multiple links between source and destination. This irregular load also affects UAVs' battery consumption, necessitating an adequate solution to meet these challenges fully. This paper proposes an SDN-based framework for UAV elements that monitors frequent changes in the network topology. Based on this monitoring, an algorithm is designed which distributes traffic load evenly on different links of multi-UAV systems. UAV networks have limited resources; therefore, battery limitations are also considered, and traffic is shifted to a path where elements have more battery. Moreover, a flight control mechanism is proposed to avoid collisions due to the high dynamics of UAVs. Extensive simulation results show that the traffic load is distributed evenly on multiple links connecting different systems with less battery consumption.