Experimental Results Analyzes in Resilient Mechanism for SDN-Based UAV Network Applied to Environmental Protection Area Surveillance

Abstract

A surveillance system requires repetitive and uninterrupted actions, typically related to large extension places, and difficult access. In this context, a multiple Unmanned Aerial Vehicles (multi-UAV) system is a good alternative for overcoming the requirements imposed by this application. With this in mind, the UAVs have to work cooperatively and exchange information to finish the mission. However, managing and keeping the communication between UAVs is a challenge that has been investigated. So, the SD-FANET is an SDN architecture developed to mitigate this communication problem. SD-FANET has a hierarchical distributed control plane that provides a resilience mechanism to overcome failures during a mission runtime. The three-step strategy (detection, election, and recovery) allows the control plane outperforms failures and works continuously as long as there are nodes in the UAV network. Experimental tests were performed in three scenarios. In all of them, the controller executed the resilience mechanism and keep going working. The mean time was 1,94 seconds to 300 executions. The PDF of results was similar to a normal distribution demonstrating the behavior of the recovery time.