Modern embedded computing systems tend to be heterogeneous in the sense of being composed subsystems with very different characteristics, which communicate and interact a variety ways-synchronous or asynchronous, buffered unbuffered, etc. Obviously, when designing such systems, modeling language needs reflect this heterogeneity. Today's environments usually offer variant what we call amorphous heterogeneity address problem. This paper argues that manner leads unexpected hard-to-analyze...

To achieve good performance in systems with limited computer resources, the constraints of implementation platform must be taken into account at design time. facilitate this, software tools are needed to analyze and simulate how timing affects control performance. This article describes two such tools: Jitterbug TrueTime.

The paper presents the emerging field of integrated control and CPU-time scheduling, where more general scheduling models methods that better suit needs systems are developed. This creates possibilities for dynamic flexible frameworks, design methodology takes availability computing resources into account during allows online trade-offs between performance resource utilization.

High-performance embedded systems require the execution of many applications on multicore platforms and are subject to stringent restrictions constraints. The ACTORS project approach provides temporal isolation through resource reservation over a platform, adapting available resources basis overall quality requirements. architecture is fully operational both ARM MPCore x86 platforms.

This paper presents the OpenDF framework and recalls that dataflow programming was once invented to address problem of parallel computing. We discuss problems with an imperative style, von Neumann programs, present what we believe are advantages using a model. The CAL actor language is briefly presented its role in ISO/MPEG standard discussed. Dataflow Interchange Format (DIF) related tools can be used for analysis actors networks, demonstrating approach. Finally, overview case study...

The complexity and dynamicity of microservice architectures in cloud environments present substantial challenges to the reliability availability services built on these architectures. Therefore, effective anomaly detection is crucial prevent impending failures resolve them promptly. Distributed data analysis techniques based machine learning (ML) have recently gained attention detecting anomalies systems. ML-based mostly require centralized collection processing, which may raise scalability...

The paper presents a Matlab toolbox for simulation of real-time control systems. basic idea is to simulate kernel in parallel with continuous plant dynamics. allows the user explore timely behavior algorithms, and study interaction between tasks scheduler. From research perspective, it also becomes possible experiment more flexible approaches systems, such as feedback scheduling. importance unified approach design systems discussed. implementation described some detail number examples are given.

The paper presents a feedback scheduling mechanism in the context of co-design scheduler and control tasks. We are particularly interested controllers where execution time may change abruptly between different modes, such as hybrid controllers. proposed solution attempts to keep CPU utilization at high level, avoid overload, distribute computing resources evenly among is implemented periodic or sporadic task that assigns sampling periods based on execution-time measurements. also communicate...

The paper presents a computational model for real-time control tasks, with the primary goal of simplifying and scheduling codesign problem. combines time-triggered I/O inter-task communication dynamic, reservation-based task scheduling. To facilitate short input-output latencies, may be divided into several segments. Jitter is reduced by allowing only at beginning end segment. A key property that both schedulability performance will depend on reserved utilization factor only. This enables...

The paper presents the control server, a real-time scheduling mechanism tailored to and signal processing applications. A server creates abstraction of task with specified period fixed input-output latency shorter than period. Individual tasks can be combined into more complex components without loss their individual guaranteed fixed-latency properties. I/O occurs at predefined points in time, which inputs are read or controller outputs become visible. model is especially suited for codesign...

Complex control systems are heterogeneous, in the sense of discrete computer-based controllers interacting with continuous physical plants, regular data sampling interleaving irregular communication and user interaction, multilayer multimode laws. This heterogeneity imposes great challenges for system design terms end-to-end performance modeling simulation, traceable refinements from algorithms to software/hardware implementation, component reuse. paper presents an actor-oriented methodology...

Hybrid controllers for a double-tank system and heating/ventilation process have been designed implemented. Both simulations real experiments are presented. It is shown that hybrid controller consisting of time-optimal combined with PID gives very good performance. The easy to implement. gives, in one its forms, guaranteed closed loop stability.

A hybrid control system consisting of a PID controller, time-optimal controller and switching strategy is investigated. specific strategy, guaranteeing stability, tested. To facilitate the real time implementation linear approximations optimal curves are used.

A novel method using the cloud to implement a variable horizon model predictive controller is presented. In case of sudden long delays and downtime, graceful degradation used. Robust, best effort strategies allow industrial grade use powerful, efficient, quickly improving ecosystems. The strategy finds in, for example, non-linear control problems, proposed can be generalized robust scalable controllers that benefit from technology. We show results two selection strategies, service...

A two-tier architecture for cloud-based MPC is presented consisting of a high rate in the cloud and low on local device. The system use as nominal controller but switches to case an unresponsive network. two MPCs are designed be similar each other possible except sampling rate. Different alternatives when execute how perform switching presented. approach evaluated by simulation.

In this paper we lay out a case for the use of dataflow programming and CAL language as way addressing current challenges in parallel hardware such multicore systems FPGAs. We show how design balances conflicting concerns expressiveness, analyzability, implementability, making it promising tool implementation stream processing applications. The itself well considerations are presented illustrated with number different cases from wide range application domains.

In this paper we present a computational offloading strategy with graceful degradation for executing Model Predictive Control using the cloud. We show method which allows seamless control assistance and design of flexible controllers edge examplify cyber-physical-system at high frequency illustrate how system can be improved while keeping cost down.