Towards a new era in publishing Dedicated Systems Magazine
Martin explains his new approach in publishing: it is a concept bringing together paper, web and CD publishing. 

By Martin Timmerman, Chief-Editor of Dedicated Systems Magazine

By Tony Romero and Sean O'Brien, sr. product manager and software engineering manager.
Performance Technologies.

Search Performance Technologies on this site.

Full article version:  
Back to top

FSMLabs RTLinux PSDD: Hard real-time with memory protection.
RTLinux was designed to run at the highest possible speed with the lowest possible overhead. For most of its existence as a product, RTLinux did not offer memory protection for threads on the grounds that the overhead in terms of context switches and requests for system services was unacceptable. The rapid increase in speed and cache size of modern microprocessors, however, and requirements of customers in aerospace and manufacturing test forced a re-evaluation. After a significant design effort including several false starts, RTLinux Process Space Development Domain (PSDD) was released in 2001. The challenges were to make the system fit cleanly within the RTLinux POSIX threads API, to minimize additional overhead, to prevent the new system from interfering at all with the existing unprotected thread system, and to produce a reliable and maintainable design. PSDD was initially aimed at FSMLabs customers in aerospace manufacturing test. Several of these customers use test frameworks that mix-and-match software simulations and control programs for physical components. Since simulation codes are often very complex and come from multiple suppliers, it is necessary to ensure that coding errors in any simulation component cannot result in undetected changes to data or code of other software - especially software controlling large and powerful physical components. Memory protection is necessary but not in itself a complete solution. PSDD tools include methods for incorporating Fortran and C++ simulations in control loops, a secondary frame or slot scheduler, full support for SMP, and protection against scheduling overruns.

By Victor Yodaiken & Michael Barabanov
FSMLabs Inc.

Search FSMLabs on this site.

Full article version:
Back to top

ProcessLogix - SCADA for Process Productions
In present time not any SCADA-system represented in the market of IT-technologies, can't "to be
praised" by dynamic self-configuring of System. In the article is described the modern SCADA-system of the leading manufacturer of means and systems for automation of technological processes of company Honeywell - ProcessLogix Release 300.1. As was writed in the previous article, ProcessLogix is a hybrid of hardware-software of the world-leaders in the field of
automation - Honeywell and Rockwell Automation. ProcessLogix has a high degree of reserved (controllers, servers, network low and high level), high degree of fault tolerance, selfdiagnosis open
DataBase, client-server architecture. One from distinctive features ProcessLogix from SCADA-systems, represented in the world market, is dynamic self-configuration. In real-time configuring all System, in particulary:
*Trends - set of any parameters in a trend (up to 8 parameters), ranging of map, setup interval of registration.
*Database of parameters - for registration in trends (for example, the parameter was circumscribed for registration with an interval 1 minute - Standard History and it is necessary for registering with an interval per 5 seconds.
*Reports - select the necessary period, level and reports is selected.
*Displays - in the editor displays HMI is draw the mnemonic scheme and saved in the necessary directory. This display automatically occurs in the list of displays HMI. *Events - nothing to configure! Everything, that hits in alarm and message, is registered automatically. Operations of an operator - modification settings and coefficients of the PID-instruction, modification of the user, emergency condition of a system (absence of connection with controllers, shutdown and reboot of servers) - all is registered by system automatically!
This some properties of a system, which gives advantage of use in process productions, such as oil processing. Don't to compile the project for to add mnemonic scheme or to change a configuration of a trend or to add new trend, to change parameters of automatic formed of the report or to change of set-up of a System.

By Roman Krutskikh, Control System Engineer, Bachelor of Engineering and Technologies,
JSC "NizhnevartovskASUneft"
ProcessLogix Inc.

Search ProcessLogix on this site.

Full article version:
Back to top

A flow control approach and interleaving method for Real-Time application in high-speed network.
Nowadays, many multimedia and real time applications are more and more greedy in
term of bandwidth and require a technology which meets this need. In this respect, the
Asynchronous Transfer Mode ATM is widely used as a transfer technique for such applications [Djemal 2000]. It consists of a connected oriented network, where data is carried in a short fixed-length packet of 53 octets, called cell. This technology was initially
designed to permit simultaneous carrying of various services all supported equivalently without any knowledge of its cell content, i.e. without discriminating some cells against others. Among all the services, variable bit rate in multimedia environment and especially digital video communication should contribute as an important fraction of the overall traffic. Due to the nature of the compression technique, the MPEG-2 sequence is inherently variable asynchronous, stochastic and non stationary. In order to provide an agreed quality of service, the network interface must be able to ensure minimal congestion conditions and to optimize the bandwidth allocation according to the need of the real time constraints.
On the other hand, the residential audio visual applications such as TV broadcasting, and Video On
Demand (VOD) will works on the basis of high speed network techniques. These video applications will widely use the motion picture expert group (MPEG) compression standards to save network resources. These video compression algorithms are mostly based on efficient DCT and block oriented techniques DCT, H.261 (video codec for audiovisual services at Px64 Kbps), ISO JPEG (Joint Picture Expert Group) and ISO MPEG-1 (Moving Picture Expert Group) [Schaphorst 96]. Unfortunately, they have been designed to support specific services and applications, and they can not interoperate or run without restriction over based ATM network. For instance, the H.261 scheme was initially designed to operate over a pecific digital networks like Narrowband-ISDN with a low rate about Px64Kbit/s to carry on teleconferencing applications where motion is naturally more limited. In the same way, JPEG algorithm only emphasizes images and do not apply any inter-coding algorithms to take into account temporal redundancy of moving picture sequences. For this application, processing delay was not a major concern for both H.261 and JPEG-1 [ISO/IEC DIS 2-1172 93].
Until now, few works have considered to allow a new higher quality video coding over the B-ISDN network related to the application which is a delay sensitive [Mehaoua 96,Seckin 2000]. Consequently, an important effort should be devoted to inter-operate a heterogeneous standard like MPEG-2 [ISO/IEC 13818-2 96] and ATM network to benefit from this environment for real time applications with agreed quality of service [Frater 99]. In fact, the transport of Picture compressed
video applications over ATM network introduces several issues that must be considered in order to ensure a high end-to-end quality. This includes the choice of adaptation layer packets for the network side, the traffic management mechanism, the cell mapping algorithm, and Interleaving and recovery schemes.

By By Belgacem Bouallegue, Ridha Djemal, Hattab Guesmi, Rached Tourki from Laboratoire Microélectronique (Tunisia) and Jean -Philippe Diguet from Laboratoire Lester (France).

Search Laboratoire on this site.

Full article version:
Back to top