Ricardo Marau

Enhancing real-time communication over cots ethernet switches

Switched Ethernet arose in the last decade as a means to increase global throughput with parallel switching paths, segment the network and create isolated collision domains, thus reducing the non-determinism of the original shared Ethernet. However, COTS Ethernet switches still suffer from a few drawbacks that affect negatively their real-time communication capabilities. For example, there can be overflows in ports queues with consequences across ports, priority levels and virtual LANs, and the number of priorities is too short for any kind of priority-based scheduling. Moreover, switches present extra latencies and jitter due to the need to interpret frame addresses and also due to different internal architectural solutions. In this paper we propose using the Flexible Time-Triggered communication paradigm to enhance the temporal behavior of Ethernet switches with respect to periodic streams. We explain the system architecture and we present a formulation of the global periodic traffic scheduling problem handled by the FTT master. Simulation and experimental results show the advantages of using such synchronized framework.

Online QoS Management for Multimedia Real-Time Transmission in Industrial Networks

A growing number of industrial applications incorporate multimedia information processing. These multimedia applications are commonly distributed and subject to time constraints that must be met across networks without creating intolerable interference over typical control flows. However, multimedia traffic, in general, and video streaming, in particular, have specific characteristics that conflict with the operational framework of conventional real-time protocols. In particular, video compressors generate highly variable bit-rate streams that mismatch the constant-bit-rate channels typically provided by real-time protocols, severely reducing the efficiency of network utilization. This paper focuses on low-latency multimedia transmission over Ethernet with dynamic quality-of-service (QoS) management. We propose a multidimensional mechanism that controls, in an integrated way, both the compression parameters and the network bandwidth allocated to each stream. The goal is to provide the best possible QoS to each stream, recomputing the compression levels and network bandwidth whenever significant events, such as channel setup/teardown, or structural changes happen. This paper also presents novel QoS metrics based both on the image quality and network parameters. Several experiments with prerecorded video streams illustrate the advantages of the proposed approach and the convenience of the metrics.

Performing Flexible Control on Low-Cost Microcontrollers Using a Minimal Real-Time Kernel

In recent years, approaches to control performance and resource optimization for embedded control systems have been receiving increased attention. Most of them focus on theory, whereas practical aspects are omitted. Theoretical advances demand flexible real-time kernel support for multitasking and preemption, thus requiring more sophisticated and expensive software/hardware solutions. On the other hand, embedded control systems often have cost constraints related with mass production and strong industrial competition, thus demanding low-cost solutions. In this paper, it is shown that these conflicting demands can be softened and that a compromise solution can be reached. We advocate that recent research results on optimal resource management for control tasks can be implemented on simple multitasking preemptive real-time kernels targeting low-cost microprocessors, which can be easily built in-house and tailored to actual application needs. The experimental evaluation shows that significant control performance improvement can be achieved without increasing hardware costs.

Utilization-based schedulability analysis for switched Ethernet aiming dynamic QoS management

Ethernet switches are typically found in many large-scale distributed real-time systems providing low-end transactions as well as bulk backbone routing to real-time applications. The FTT-SE protocol (Flexible Time-Triggered communication over Switched Ethernet) is a recent proposal to bypass the limitations of conventional switches in terms of real-time behavior while catering for growing requirements on dynamic reconfigurability and adaptability. For this end, this paper develops linear time-complexity and memory-efficient on-line admission control tests based on utilization bounds for Rate-Monotonic and EDF scheduling on Ethernet switches using FTT-SE, which are suited for dynamic Quality of Service (QoS) management. Our analysis also has broader applicability in general periodic task sets with bounded release delays. For FTT-SE with 100Mbps links and 1500 bytes of maximum packet size, our sufficient schedulability condition achieves an utilization bound of 61% for RMS and 88% for EDF. Simulation results on randomly generated task sets demonstrate that such bounds are within 18% and 5% utilization of the ideal tests for RMS and EDF, respectively.

Implementing a distributed sensing and actuation system: The CAMBADA robots case study

The use of distributed computing architectures has become commonplace in complex embedded systems with potential advantages, for example, in terms of scalability, dependability and maintainability. One particular area in which that trend can be witnessed is mobile autonomous robotics in which several sensors and actuators are interconnected by means of a control network. In this paper we address one case study concerning the CAMBADA robots that were developed at the University of Aveiro for the Robocup Middle Size League. These robots have a distributed architecture with two layers, a coordination layer responsible for the global behaviors and a distributed sensing and actuating layer that conveys internal state information and executes coordination commands. This paper focuses on the latter layer, which is based on the FTT-CAN protocol, following a network-centric approach that provides an efficient framework for the synchronization of all systems activities. We describe the computing and communication requirements, the robot architecture, the system design and implementation, and finally we provide experimental results that show advantages with respect to a non-synchronized distributed approach

Flexible, efficient and robust real-time communication with server-based Ethernet Switching

The information exchanged in Networked Embedded Systems is steadily increasing in quantity, size, complexity and heterogeneity, with growing requirements for arbitrary arrival patterns and guaranteed QoS. One of the networking protocols that is becoming more common in such systems is Ethernet and its real-time Ethernet variants. However, they hardly support all the referred requirements in an efficient manner since they either favour determinism or throughput, but not both. A potential solution recently proposed by the authors is the Server-SE protocol that uses servers to confine traffic associated to specific applications or subsystems. Such an approach is dynamically reconfigurable and adaptive, being more bandwidth efficient while providing composability in the time domain. This paper proposes integrating the servers inside the Ethernet switch, boosting both the flexibility and the robustness of Server-SE, allowing, for example, the seamless connection of any Ethernet node. The switch is an FTT-enabled Ethernet Switch and the paper discusses two specific ways of integrating the servers, namely in software or in hardware. These options are described and compared analytically and experimentally. The former favours flexibility in the servers design and management while the latter provides lower latency.

Designing a costumized Ethernet switch for safe hard real-time communication

The use of switched Ethernet for precise and safe real-time communication still suffers from undesired phenomena that range from the blocking caused by long non-preemptive frames to lack of protection against errors in the time domain and also couplings across virtual LANs and even priority levels via internal switch shared resources. In this paper we propose a novel switch architecture enhanced with resource reservation mechanisms, based on the Flexible Time-Triggered paradigm, which enforces strict service differentiation, blocking-free forwarding and timing errors confinement. Experimental results of a preliminary 4-port prototype based on an FPGA validate the desired properties and exhibit the potential of the enhanced Ethernet switch.

A synthesizable ethernet switch with enhanced real-time features

The use of switched Ethernet for safe real-time communication still suffers from undesired phenomena, such as blocking caused by long non-preemptive frames, lack of protection against errors in the time domain, couplings across virtual LANs and priority levels via internal switch shared resources. Recently, a few solutions were proposed to cope with such phenomena. One such solution is based on an enhanced switch following the Flexible Time-Triggered paradigm, which enforces strict service differentiation with any kind of traffic scheduling, blocking-free forwarding and timing errors confinement. In this paper we propose a new architecture following an hardwaresoftware co-design approach that simplifies the development of the enhanced switch features by detaching the traffic scheduling from the traffic switching. The paper shows experimental results with an actual switch prototype that confirm the desired switch properties.

A middleware to support dynamic reconfiguration of real-time networks

The middleware is an important software component when designing an application, facilitating the development and deployment of the applications. In the case of Distributed Embedded Systems (DES), the middleware should provide basic functionalities to abstract the complexity that results from network distribution, namely data consistency, events synchronization and resource management. Often DES applications exhibit real-time requirements and have to deal with dynamic environments that present evolving requirements. While some middleware architectures have been proposed to address resource provisioning and QoS management, none of those middlewares supports dynamic resource reconfiguration while providing real-time guarantees. This paper proposes a middleware layer, based on the services provided by a flexible real-time communication protocol, addressing distribution abstraction, dynamic reconfiguration and dynamic QoS management under real-time constraints.

Controlling multi-switch networks for prompt reconfiguration

Recent trends in distributed embedded systems, such as those found in avionics and trains, have shown an increase in the amount and heterogeneity of the information that needs to be exchanged, together with a growing importance of supporting dynamic reconfiguration and adaptive behaviors. In this paper we focus on Ethernet technologies with real-time reconfiguration support and we address the case of middle-size networking infrastructures with a few switches. We use the FTT-SE protocol with the needed adaptations to support dynamic heterogeneous real-time transactions in multi-hop networks. The paper presents a worst-case response-time analysis that provides timeliness guarantees, improving the results obtained with another previous analysis, decreasing the needed network capacity for guaranteed schedulability by 25% on average. Practical experiments and simulation results validate the proposed approach and analysis.