Distributing Synchronous Programs on a Loosely Synchronous, Distributed Architecture

نویسندگان

  • ALBERT BENVENISTE
  • Albert Benveniste
چکیده

In this short note we discuss the distribution of synchronous programs on distributed, asynchronous architecture of the type preferred for embedded real-time control applications. In such architectures, no global clock is broadcast. Instead, the bus, as well as its users are triggered by local periodic clocks. To improve fault-tolerance and robust-ness, these clocks are not synchronized, neither physically nor logically. Therefore such an architecture has some (restricted) degree of asynchrony, we call it loosely synchronous. A typical example of such architecture is the osek protocol, used for embedded electronics in automotive industry. In this note we show how synchronous programs can be safely distributed on loosely synchronous architectures. The method described here may not be novel to some communities, but it aims at ensuring awareness of the community of engineers in the area of real-time, embedded control systems. Distribution de programmes synchrones sur une architecture mollement synchrone R esum e : Cette note pr esente une approche tr es simple pour la r epartition de programmes synchrones sur des architectures typiques des bus de terrain en automatique temps-r eel (un exemple connu est osek dans le domaine de l' electronique embarqu ee en automobile). De telles architectures peuvent ^ etre vues comme un bus dot e d'une horloge p eriodique. Les usagers du bus proc edent egalement a des acc es p eriodiques. Le point important est que 1/ toutes ces horloges sont purement locales et non synchronis ees, et 2/ rien ne doit ^ etre bloquant. Il s'agit donc d'architectures r eparties qui ne satisfont pas au mod ele des langages synchrones, car elle n'assure pas d'horloge logique globale. Nous qualiions ces architectures de mollement synchrones. Dans cette note, nous proposons donc une approche tr es simple pour r epartir, de mani ere correcte par construction, des programmes synchrones sur une architecture mollement syn-chrone.

برای دانلود رایگان متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

ثبت نام

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

منابع مشابه

Diffusive Load Balancing of Loosely-Synchronous Parallel Programs over Peer-to-Peer Networks

The use of under-utilized Internet resources is widely recognized as a viable form of high performance computing. Sustained processing power of roughly 40T FLOPS using 4 million volunteered Internet hosts has been reported for embarrassingly parallel problems. At the same time, peer-to-peer (P2P) file sharing networks, with more than 50 million participants, have demonstrated the capacity for s...

متن کامل

Distributing Synchronous Programs Using Bounded Queues, a coordinated traffic signal application

This paper is about the modular compilation and distribution of a sub-class of Simulink programs [10] across networks using bounded FIFO queues. The problem is first addressed mathematically. Then, based on these formal results, a software library for the modular compilation and distribution of Simulink program is given. The performance the library is given. The value of synchronous programming...

متن کامل

A Loosely Time-Triggered Architectures: Improvements and Comparisons

Loosely Time-Triggered Architectures (LTTAs) are a proposal for constructing distributed embedded control systems. They build on the quasi-periodic architecture, where computing units execute nearly periodically, by adding a thin layer of middleware that facilitates the implementation of synchronous applications. In this paper, we show how the deployment of a synchronous application on a quasi-...

متن کامل

How to scale distributed deep learning?

Training time on large datasets for deep neural networks is the principal workflow bottleneck in a number of important applications of deep learning, such as object classification and detection in automatic driver assistance systems (ADAS). To minimize training time, the training of a deep neural network must be scaled beyond a single machine to as many machines as possible by distributing the ...

متن کامل

Programming Research Group Conservative Discrete-event Simulations on Bulk Synchronous Parallel Architectures Conservative Discrete-event Simulations on Bulk Synchronous Parallel Architectures

All the parallel discrete-event simulation algorithms developed so far have been designed to suit a speciic parallel model (e.g., a PRAM model, a MP-RAM model, etc.). This paper presents several versions of conservative parallel discrete-event simulation algorithms developed around a unifying model for general purpose parallel computer design and programming, namely around the Bulk Synchronous ...

متن کامل

ذخیره در منابع من


  با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید

عنوان ژورنال:

دوره   شماره 

صفحات  -

تاریخ انتشار 1999