On Edge Coordination in Highly Dynamic Cyber-Physical Systems for Emergency Response

Mobile edge computing replicates cloud functions at the edge, thus reducing latency, improving responsiveness, and utilizing network bandwidth in resource-intensive applications such as Cyber-Physical Systems (CPS) for disaster response. These systems, which have gained significant interest from first responders, present immense potential to purvey the necessary computational power in highly dynamic and volatile environments like disaster response. However, existing edge computing solutions operate under the assumption that devices at the edge remain connected through a reliable edge network. While such systems bring enormous possibilities to first responders through new emergency response applications on handheld devices, they still suffer from poor inter-connectivity and, thus, remain ineffective. In this paper, we explore the impact of unreliable edge connectivity on edge computing functions, by deploying and evaluating a highly dynamic edge computing CPS consisting of an unmanned aerial vehicle (UAV) and mobile edge computing devices. We present our results from the deployment, showing limitations of current state-of-the-art solutions in tackling the challenges posed by unpredictable connectivity. We propose a solution for discovering services over intermittent LTE connections, to address some of the issues identified in the deployment.