Analysis of a Queueing Model for Payload Data Processing in a UAV Remote Sensing Environment

Abstract

UAV platforms are showing renewed interest in the field of remote sensing, leveraged by their high maneuverability and reduced preparation time. However, the on-board data processing is still a challenge. Typical UAV missions record data and are then processed after the flight, increasing the data delivery from days to weeks. This is prohibitive in situations where real-time data analysis is required for fast decision making in remote sensing and SAR missions. With the advancement of embedded computing systems and multi-core architectures, remote sensing can benefit from online data processing and faster result delivery. There, however, remains a challenge for system designers to use hardware resources efficiently. This study aims to explore an existing data processing software architecture using queuing model approach and proposes a better architecture with reduced latency and increased throughput.