| Abstract | Unmanned Aerial Vehicles (UAVs) navigate by observing their surroundings using on-board cameras and sensors. UAVs also collect data using their onboard devices, compute the data, and analyze the information to decide which direction to move next. Onboard computation and processing are known to have a significant overhead on UAVs and might limit the time of flight, speed and efficiency of UAVs. Thus, the idea of offloading computation and processing to servers that are available at the edge of the network and close to the UAVs has been strongly advocated and researched recently. This paper aims to leverage offloading UAV computations to the edge while exploring different implementations and approaches of computation offloading. To achieve the aim, we reviewed the papers from IEEE conferences and other verified institutions such as NASA and Cornell University and did a literature survey based on our findings. Our results implicate the overview of computation offloading and its implementations for various UAV applications. This study emphasizes on various offloading approaches for UAV computations to Mobile Edge Computing (MEC) servers along with their requirements, advantages, and disadvantages. |
