Search

Engine control applications include tasks with variable rate dependent behaviour (VRB-tasks) that adapt their behavior as a function of the engine speed. For such tasks, the frequency of activation, execution times and deadlines depend on the rotation speed of the engine. In recent years exact methods for the analysis of such VRB-tasks have been introduced, where the task is represented by the accumulated execution requirement in relation to the time interval. In this paper we identify the periodic behaviour of this curve and determine the recurrent term. The result is the first method to have precise knowledge of the exact execution requirement of VRB-tasks for arbitrary high intervals.

Offset dependencies between tasks/messages ensure that there exists a fixed time interval between the activation of a set of tasks/ messages, thereby avoiding simultaneous activations. Offset-based analysis techniques include time offsets into the computational model and perform response time analysis by taking into account the time offsets. However, for systems with task offsets where tasks are not released simultaneously, identifying the critical instant is not straight-forward. It leads to a combinatorial problem where several busy window periods have to be explored to identify the critical instant of the task, thereby increasing the analysis run-time exponentially. In this paper, we propose a novel heuristics and a flexible approximation technique to limit the critical arrival pattern of tasks and reduce the run-time of an offset-based response time analysis., acceptedVersion

Underwater communication with autonomous underwater vehicles (AUVs) has strong demands on the modems caused by the constantly changing signal propagation (multipath propagation, scattering, diffraction and refraction at thermal layers, etc.) of the underwater channel. These demands typically lead to a modem designed to match specific conditions. This results in two major problems, first we do not want to develop for each changing environment a specific hardware and second we need a robust communication to distribute necessary information about the state of the system, current condition of the environment or to coordinate group missions. In this paper we introduced a approach to serve booth of the problems. We implemented with a flexible generic design a robust underwater communication system. The paper introduce a designed approach for a generic reusable acoustic underwater communication based on a pipelined data stream processing system on a Field Programmable Gate Array (FPGA). Exploiting the inherent synchronization mechanism of an hand-shake-based streaming bus interface to form arbitrary long pipelines composed of detached and reusable modules representing the coding and modulation algorithms without the need of an external coordination mechanism. We show that our approach is feasible, matches the problems and can easily be changed or extended to a Multiple Input Multiple Output (MIMO)-system while the recommendation on energy consumption is nearly equally and area is reduced by 45% compared to other system implementation.

In automotive production car engines there are tasks, where the frequency of activation and execution times vary with the angular velocity of the engine. The variable rate-dependent behavior (VRB) task model has been proposed as a means of modeling this behavior. In the literature a variety of real-time analysis have been developed for tasks with this dependency of an angular velocity. However, these analysis focus on mono-processor systems. In this paper we propose the first analysis for distributed systems including VRB-tasks.