Architecture and Execution Model for Quadcopter Control

Abstract

This paper proposes a temporal model for executing the computational load of a quadcopter-type unmanned aerial vehicle (UAV), treated as a partitioned real-time system of mixed criticality running on a multiprocessor platform. The analysis is carried out using \textit{End-to-End Flows} (ETEF) diagrams, which allow the interaction between tasks to be characterised and the information flow of functional loops to be modelled at a high level of abstraction. The system under consideration is structured using software partitions that group specific functionalities and their associated tasks, following principles of partitioned architectures inspired by IMA approaches and standards such as ARINC 653. The proposed approach avoids exhaustive analysis at individual task level, focusing instead on the representation of precedence and communication relationships through temporal flows. As a result, a structured description of the system’s behaviour is obtained, based on dependencies between tasks and the organisation of the various functional loops. This model provides a solid basis for subsequent temporal analysis of the system.