Hardware-Independent Self-Discovery of Peripherals and Modules of a Self-Adaptive Avionics Platform

Abstract

The latest generation of large passenger aircraft use Integrated Modular Avionics (IMA) with standardized hardware and software to host aircraft functions on a distributed platform with shared hardware devices. While IMA have reduced the size, weight, power consumption and cost (SWaP-C) of avionics systems, the configuration of IMA platforms remains a time-consuming and error-prone task. We address this shortcoming with our Plug&Fly self-adaptive avionics platform approach, which is intended to provide self-* properties of Autonomic Computing such as self-configuration, self-optimization, self-healing, and self-protection. In order to achieve a continuous awareness of the present hardware topology within the distributed platform for the self-configuration, we present a hardware-independent model-based topology discovery in this paper. Our approach consists of four building blocks: (1) a generic topology discovery algorithm solely reliant on communication primitives, (2) a hardware and communication abstraction layer, (3) a message serialization and self-description, and (4) a run-time model data aggregator. We demonstrate our self-discovery approach on a tethered octocopter running a flight control law on heterogeneous computing hardware with Ethernet and CAN networks.