Self-Adaptive Embedded Network

Paper #:
  • 2015-01-2523

  • 2015-09-15
Coustal, P. and Tailliez, F., "Self-Adaptive Embedded Network," SAE Technical Paper 2015-01-2523, 2015,
In the Integrated Modular Avionics (IMA) domain, THALES developed a high performance communication network named SAEN (Self Adaptive Embedded Network).SAEN is a switchless network solution, fully embedded in a single Network Component Interface (NCI), aimed to interconnect easily several modules of a system, in any mesh network topology. Once each module is equipped with its network component, just connect them together to realize the wanted topology and switch ‘on’ the modules power supplies. At power-on, all the nodes of the network aggregate to form a complete global and coherent network, autonomously managing its configuration and the optimal static routing between any emitter and receiver. The constituted network is deterministic, autonomous, self-discovering, and auto-adapting to the network variations and guarantees an optimal routing in any situation of the graph, as long as a path exists. The interest of managing mesh topology resides in the intrinsic robustness offered by the graph connectivity.This solution is being implemented in the new mission computer embedded in the latest French combat aircraft. For this application, the components drive a 2 Gigabits per second (Gbps) physical layer on a copper backplane. It is to be noticed that for this application, the network is highly constrained since the mission computer is likely to reallocate dynamically the applications on various processing modules.The paper describes the basic elements of graph theory that allow to explain the robustness properties linked to the graph connectivity. Then some linear algebra used to compute the optimal routing and its optimization in a wired logic module of the component. Finally, the paper presents a brief description of the application on the aircraft mission computer.
SAE MOBILUS Subscriber? You may already have access.
Members save up to 42% off list price.
HTML for Linking to Page
Page URL

Related Items