ACODS: adaptive computation offloading for drone surveillance system

Unmanned Aerial Vehicle (UAV) surveillance systems allow for highly advanced and safe surveillance of hazardous locations. Further, multi-purpose drones can be widely deployed for not only gathering information but also analyzing the situation from sensed data. However, mobile drone systems have limited computing resources and battery power which makes it a challenge to use these systems for long periods of time or in fully autonomous modes. In this paper, we propose an Adaptive Computation Offloading Drone System (ACODS) architecture with reliable communication for increasing drone operating time. We design not only the response time prediction module for mission critical task offloading decision but also task offloading management module via the Multipath TCP (MPTCP). Through performance evaluation via our prototype implementation, we show that the proposed algorithm achieves significant increase in drone operation time and significantly reduces the response time.