Supporting intra-task parallelism in real-time multiprocessor systems

Os sistemas de tempo real modernos geram, cada vez mais, cargas computacionais pesadas e dinâmicas, começando-se a tornar pouco expectável que sejam implementados em sistemas uniprocessador. Na verdade, a mudança de sistemas com um único processador para sistemas multi- processador pode ser vista, tanto no domínio geral, como no de sistemas embebidos, como uma forma eficiente, em termos energéticos, de melhorar a performance das aplicações. Simultaneamente, a proliferação das plataformas multi-processador transformaram a programação paralela num tópico de elevado interesse, levando o paralelismo dinâmico a ganhar rapidamente popularidade como um modelo de programação. A ideia, por detrás deste modelo, é encorajar os programadores a exporem todas as oportunidades de paralelismo através da simples indicação de potenciais regiões paralelas dentro das aplicações. Todas estas anotações são encaradas pelo sistema unicamente como sugestões, podendo estas serem ignoradas e substituídas, por construtores sequenciais equivalentes, pela própria linguagem. Assim, o modo como a computação é na realidade subdividida, e mapeada nos vários processadores, é da responsabilidade do compilador e do sistema computacional subjacente. Ao retirar este fardo do programador, a complexidade da programação é consideravelmente reduzida, o que normalmente se traduz num aumento de produtividade. Todavia, se o mecanismo de escalonamento subjacente não for simples e rápido, de modo a manter o overhead geral em níveis reduzidos, os benefícios da geração de um paralelismo com uma granularidade tão fina serão meramente hipotéticos. Nesta perspetiva de escalonamento, os algoritmos que empregam uma política de workstealing são cada vez mais populares, com uma eficiência comprovada em termos de tempo, espaço e necessidades de comunicação. Contudo, estes algoritmos não contemplam restrições temporais, nem outra qualquer forma de atribuição de prioridades às tarefas, o que impossibilita que sejam diretamente aplicados a sistemas de tempo real. Além disso, são tradicionalmente implementados no runtime da linguagem, criando assim um sistema de escalonamento com dois níveis, onde a previsibilidade, essencial a um sistema de tempo real, não pode ser assegurada. Nesta tese, é descrita a forma como a abordagem de work-stealing pode ser resenhada para cumprir os requisitos de tempo real, mantendo, ao mesmo tempo, os seus princípios fundamentais que tão bons resultados têm demonstrado. Muito resumidamente, a única fila de gestão de processos convencional (deque) é substituída por uma fila de deques, ordenada de forma crescente por prioridade das tarefas. De seguida, aplicamos por cima o conhecido algoritmo de escalonamento dinâmico G-EDF, misturamos as regras de ambos, e assim nasce a nossa proposta: o algoritmo de escalonamento RTWS. Tirando partido da modularidade oferecida pelo escalonador do Linux, o RTWS é adicionado como uma nova classe de escalonamento, de forma a avaliar na prática se o algoritmo proposto é viável, ou seja, se garante a eficiência e escalonabilidade desejadas. Modificar o núcleo do Linux é uma tarefa complicada, devido à complexidade das suas funções internas e às fortes interdependências entre os vários subsistemas. Não obstante, um dos objetivos desta tese era ter a certeza que o RTWS é mais do que um conceito interessante. Assim, uma parte significativa deste documento é dedicada à discussão sobre a implementação do RTWS e à exposição de situações problemáticas, muitas delas não consideradas em teoria, como é o caso do desfasamento entre vários mecanismo de sincronização. Os resultados experimentais mostram que o RTWS, em comparação com outro trabalho prático de escalonamento dinâmico de tarefas com restrições temporais, reduz significativamente o overhead de escalonamento através de um controlo de migrações, e mudanças de contexto, eficiente e escalável (pelo menos até 8 CPUs), ao mesmo tempo que alcança um bom balanceamento dinâmico da carga do sistema, até mesmo de uma forma não custosa. Contudo, durante a avaliação realizada foi detetada uma falha na implementação do RTWS, pela forma como facilmente desiste de roubar trabalho, o que origina períodos de inatividade, no CPU em questão, quando a utilização geral do sistema é baixa. Embora o trabalho realizado se tenha focado em manter o custo de escalonamento baixo e em alcançar boa localidade dos dados, a escalonabilidade do sistema nunca foi negligenciada. Na verdade, o algoritmo de escalonamento proposto provou ser bastante robusto, não falhando qualquer meta temporal nas experiências realizadas. Portanto, podemos afirmar que alguma inversão de prioridades, causada pela sub-política de roubo BAS, não compromete os objetivos de escalonabilidade, e até ajuda a reduzir a contenção nas estruturas de dados. Mesmo assim, o RTWS também suporta uma sub-política de roubo determinística: PAS. A avaliação experimental, porém, não ajudou a ter uma noção clara do impacto de uma e de outra. No entanto, de uma maneira geral, podemos concluir que o RTWS é uma solução promissora para um escalonamento eficiente de tarefas paralelas com restrições temporais.

Soil remediation time to achieve clean-up goals I: influence of soil water content

The current models are not simple enough to allow a quick estimation of the remediation time. This work reports the development of an easy and relatively rapid procedure for the forecasting of the remediation time using vapour extraction. Sandy soils contaminated with cyclohexane and prepared with different water contents were studied. The remediation times estimated through the mathematical fitting of experimental results were compared with those of real soils. The main objectives were: (i) to predict, through a simple mathematical fitting, the remediation time of soils with water contents different from those used in the experiments; (ii) to analyse the influence of soil water content on the: (ii1) remediation time; (ii2) remediation efficiency; and (ii3) distribution of contaminants in the different phases present into the soil matrix after the remediation process. For sandy soils with negligible contents of clay and natural organic matter, artificially contaminated with cyclohexane before vapour extraction, it was concluded that (i) if the soil water content belonged to the range considered in the experiments with the prepared soils, then the remediation time of real soils of similar characteristics could be successfully predicted, with relative differences not higher than 10%, through a simple mathematical fitting of experimental results; (ii) increasing soil water content from 0% to 6% had the following consequences: (ii1) increased remediation time (1.8–4.9 h, respectively); (ii2) decreased remediation efficiency (99–97%, respectively); and (ii3) decreased the amount of contaminant adsorbed onto the soil and in the non-aqueous liquid phase, thus increasing the amount of contaminant in the aqueous and gaseous phases.

Soil remediation time to achieve clean-up goals II: influence of natural organic matter and water contents

This work reports a relatively rapid procedure for the forecasting of the remediation time (RT) of sandy soils contaminated with cyclohexane using vapour extraction. The RT estimated through the mathematical fitting of experimental results was compared with that of real soils. The main objectives were: (i) to predict the RT of soils with natural organic matter (NOM) and water contents different from those used in experiments; and (ii) to analyse the time and efficiency of remediation, and the distribution of contaminants into the soil matrix after the remediation process, according to the soil contents of: (ii1) NOM; and (ii2) water. For sandy soils with negligible clay contents, artificially contaminated with cyclohexane before vapour extraction, it was concluded that: (i) if the NOM and water contents belonged to the range of the prepared soils, the RT of real soils could be predicted with relative differences not higher than 12%; (ii1) the increase of NOM content from 0% to 7.5% increased the RT (1.8–13 h) and decreased the remediation efficiency (RE) (99–90%) and (ii2) the increase of soil water content from 0% to 6% increased the RT (1.8–4.9 h) and decreased the RE (99–97%). NOM increases the monolayer capacity leading to a higher sorption into the solid phase. Increasing of soil water content reduces the mass transfer coefficient between phases. Concluding, NOM and water contents influence negatively the remediation process, turning it less efficient and more time consuming, and consequently more expensive.

Effects of sulphurization time on Cu2ZnSnS4 absorbers and thin films solar cells obtained from metallic precursors

We report the results of a study of the sulphurization time effects on Cu2ZnSnS4 absorbers and thin film solar cells prepared from dc-sputtered tackedmetallic precursors. Three different time intervals, 10 min, 30min and 60 min, at maximum sulphurization temperature were considered. The effects of this parameter' change were studied both on the absorber layer properties and on the final solar cell performance. The composition, structure, morphology and thicknesses of the CZTS layers were analyzed. The electrical characterization of the absorber layer was carried out by measuring the transversal electrical resistance of the samples as a function of temperature. This study shows an increase of the conductivity activation energy from 10 meV to 54meV for increasing sulphurization time from 10min to 60min. The solar cells were built with the following structure: SLG/Mo/CZTS/CdS/i-ZnO/ZnO:Al/Ni:Al grid. Several ac response equivalent circuit models were tested to fit impedance measurements. The best results were used to extract the device series and shunt resistances and capacitances. Absorber layer's electronic properties were also determined using the Mott–Schottky method. The results show a decrease of the average acceptor doping density and built-in voltage, from 2.0 1017 cm−3 to 6.5 1015 cm−3 and from 0.71 V to 0.51 V, respectively, with increasing sulphurization time. These results also show an increase of the depletion region width from approximately 90 nm–250 nm.

Run-time monitoring in real-time operating systems

Embedded systems are increasingly complex and dynamic, imposing progressively higher developing time and costs. Tuning a particular system for deployment is thus becoming more demanding. Furthermore when considering systems which have to adapt themselves to evolving requirements and changing service requests. In this perspective, run-time monitoring of the system behaviour becomes an important requirement, allowing to dynamically capturing the actual scheduling progress and resource utilization. For this to succeed, operating systems need to expose their internal behaviour and state, making it available to external applications, and a runtime monitoring mechanism must be available. However, such mechanism can impose a burden in the system itself if not wisely used. In this paper we explore this problem and propose a framework, which is intended to provide this run-time mechanism whilst achieving code separation, run-time efficiency and flexibility for the final developer.

The influence of inter-domain mobility on message stream response time in wired/wireless PROFIBUS-based networks

In previous works we have proposed a hybrid wired/wireless PROFIBUS solution where the interconnection between the heterogeneous media was accomplished through bridge-like devices with wireless stations being able to move between different wireless cells. Additionally, we had also proposed a worst-case timing analysis assuming that stations were stationary. In this paper we advance these previous works by proposing a worst-case timing analysis for the system’s message streams considering the effect of inter-cell mobility.

Capacity sharing and stealing in serverbased real-time systems

A dynamic scheduler that supports the coexistence of guaranteed and non-guaranteed bandwidth servers is proposed. Overloads are handled by an efficient reclaiming of residual capacities originated by early completions as well as by allowing reserved capacity stealing of non-guaranteed bandwidth servers. The proposed dynamic budget accounting mechanism ensures that at a particular time the currently executing server is using a residual capacity, its own capacity or is stealing some reserved capacity, eliminating the need of additional server states or unbounded queues. The server to which the budget accounting is going to be performed is dynamically determined at the time instant when a capacity is needed. This paper describes and evaluates the proposed scheduling algorithm, showing that it can efficiently reduce the mean tardiness of periodic jobs. The achieved results become even more significant when tasks’ computation times have a large variance.

Mechanisms for reflection-based monitoring of real- time systems

Monitoring is a very important aspect to consider when developing real-time systems. However, it is also important to consider the impact of the monitoring mechanisms in the actual application. The use of Reflection can provide a clear separation between the real-time application and the implemented monitoring mechanisms, which can be introduced (reflected) into the underlying system without changing the actual application part of the code. Nevertheless, controlling the monitoring system itself is still a topic of research. The monitoring mechanisms must contain knowledge about “how to get the information out”. Therefore, this paper presents the ongoing work to define a suitable strategy for monitoring real-time systems through the use of Reflection.

Position paper on time and event-triggered communication services in the context of e-manufacturing

Modern factories are complex systems where advances in networking and information technologies are opening new ways towards higher efficiency. Such move is being driven by market rules with ever-increasing competition levels, in search for faster time-to-market, improved process yield, non-stop operations, flexible manufacturing and tighter supply-chain coupling. All these aims present a common requirement, i.e. a realtime flow of information, from the plant-floor up to the management, maintenance, suppliers and clients, to support accurate monitoring and control of the factory. This stresses the importance achieved by the communication infrastructure in modern manufacturing industry. This paper presents the authors view concerning the current trends in modern factory communication systems. It addresses the problems of seamlessly integrating different information flows with diverse requirements, mainly in terms of timeliness. In this aspect, the debate between event-triggered and time-triggered communication is revisited as well as the joint support for both types of traffic. Finally, a view of where factory communication systems are moving to is also presented, showing the impact of open and widely available technologies.

A survey of techniques and technologies for web-based real-time interactive rendering

When exploring a virtual environment, realism depends mainly on two factors: realistic images and real-time feedback (motions, behaviour etc.). In this context, photo realism and physical validity of computer generated images required by emerging applications, such as advanced e-commerce, still impose major challenges in the area of rendering research whereas the complexity of lighting phenomena further requires powerful and predictable computing if time constraints must be attained. In this technical report we address the state-of-the-art on rendering, trying to put the focus on approaches, techniques and technologies that might enable real-time interactive web-based clientserver rendering systems. The focus is on the end-systems and not the networking technologies used to interconnect client(s) and server(s).

A framework for realistic real-time walkthroughs in a VR distributed environment

Virtual and augmented reality (VR/AR) are increasingly being used in various business scenarios and are important driving forces in technology development. However the usage of these technologies in the home environment is restricted due to several factors including lack of low-cost (from the client point of view) highperformance solutions. In this paper we present a general client/server rendering architecture based on Real-Time concepts, including support for a wide range of client platforms and applications. The idea of focusing on the real-time behaviour of all components involved in distributed IP-based VR scenarios is new and has not been addressed before, except for simple sub-solutions. This is considered as “the most significant problem with the IP environment” [1]. Thus, the most important contribution of this research will be the holistic approach, in which networking, end-systems and rendering aspects are integrated into a cost-effective infrastructure for building distributed real-time VR applications on IP-based networks.

The DEAR-COTS hard real-time subsystem

In this report, the Hard Real-Time Subsystem of DEAR-COTS is described, and the services it must provide are identified. This report is an input of ISEP/IPP and FEUP for the specification of the DEAR-COTS architecture (deliverable to the FCT).

CIDER - envisaging a COTS communication infrastructure for evolutionary dependable real-time systems

It is foreseen that future dependable real-time systems will also have to meet flexibility, adaptability and reconfigurability requirements. Considering the distributed nature of these computing systems, a communication infrastructure that permits to fulfil all those requirements is thus of major importance. Although Ethernet has been used primarily as an information network, there is a strong belief that some very recent technological advances will enable its use in dependable applications with real-time requirements. Indeed, several recently standardised mechanisms associated with Switched-Ethernet seem to be promising to enable communication infrastructures to support hard real-time, reliability and flexible distributed applications. This paper describes the motivation and the work being developed within the CIDER (Communication Infrastructure for Dependable Evolvable Real-Time Systems) project, which envisages the use of COTS Ethernet as an enabling technology for future dependable real-time systems. It is foreseen that the CIDER approach will constitute a relevant stream of research since it will bring together cutting edge research in the field of real-time and dependable distributed systems and the industrial eagerness to expand Ethernet responsabilities to support dependable real-time applications.

Ethernet goes real-time: a survey on research and technological developments

Ethernet is the most popular LAN technology. Its low price and robustness, resulting from its wide acceptance and deployment, has created an eagerness to expand its responsibilities to the factory-floor, where real-time requirements are to be fulfilled. However, it is difficult to build a real-time control network using Ethernet, because its MAC protocol, the 1-persistent CSMA/CD protocol with the BEB collision resolution algorithm, has unpredictable delay characteristics. Many anticipate that the recent technological advances in Ethernet such as the emerging Fast/Gigabit Ethernet, micro-segmentation and full-duplex operation using switches will also enable it to support time-critical applications. This technical report provides a comprehensive look at the unpredictability inherent to Ethernet and at recent technological advances towards real-time operation.

Designing real-time systems based on mono-master Profibus-DP networks

Profibus networks are widely used as the communication infrastructure for supporting distributed computer-controlled applications. Most of the times, these applications impose strict real-time requirements. Profibus-DP has gradually become the preferred Profibus application profile. It is usually implemented as a mono-master Profibus network, and is optimised for speed and efficiency. The aim of this paper is to analyse the real-time behaviour of this class of Profibus networks. Importantly, we develop a new methodology for evaluating the worst-case message response time in systems where high-priority and cyclic low-priority Profibus traffic coexist. The proposed analysis constitutes a powerful tool to guarantee prior to runtime the real-time behaviour of a distributed computer-controlled system based on a Profibus network, where the realtime traffic is supported either by high-priority or by cyclic poll Profibus messages.