Real-Time Systems. Design Principles for Distributed Embedded Applications. Herman Kopetz. Second Edition (Real-Time Systems. Design Principles for Distributed Embedded Applications. Herman Kopetz. Second Edition.pdf), страница 10

It is difficult to establish that a best-effort design operates correctly in arare-event scenario. At present, many non safety-critical real-time systems aredesigned according to the best-effort paradigm.1.5.4Resource-Adequate Versus Resource-InadequateGuaranteed response systems are based on the principle of resource adequacy, i.e.,there are enough computing resources available to handle the specified peak loadand the fault scenario. Many non safety-critical real-time system designs are basedon the principle of resource inadequacy.

It is assumed that the provision ofsufficient resources to handle every possible situation is not economically viable,and that a dynamic resource allocation strategy based on resource sharing andprobabilistic arguments about the expected load and fault scenarios is acceptable.It is expected that, in the future, there will be a paradigm shift to resourceadequate designs in many applications. The use of computers in important volumebased applications, e.g., in cars, raises both the public awareness as well as concernsabout computer-related incidents, and forces the designer to provide convincingarguments that the design functions properly under all stated conditions.

Hard realtime systems must be designed according to the guaranteed response paradigm thatrequires the availability of adequate resources.1.5.5Event-Triggered Versus Time-TriggeredThe distinction between event-triggered and time-triggered depends on the type ofinternal triggers and not the external behavior of a real-time system. A trigger is anevent that causes the start of some action in the computer, e.g., the execution of a1.6 The Real-Time Systems Market17task or the transmission of a message.

Depending on the triggering mechanisms forthe start of communication and processing actions in each node of a computersystem, two distinctly different approaches to the design of the control mechanismsof real-time computer applications can be identified, event-triggered control andtime-triggered control.In event-triggered (ET) control, all communication and processing activities areinitiated whenever a significant event other than the regular event of a clock tickoccurs. In an ET system, the signaling of significant events to the central processingunit (CPU) of a computer is realized by the well-known interrupt mechanism.ET systems require a dynamic scheduling strategy to activate the appropriatesoftware task that services the event.In a time-triggered (TT) system, all activities are initiated by the progression ofreal-time.

There is only one interrupt in each node of a distributed TT system, theperiodic real-time clock interrupt. Every communication or processing activity isinitiated at a periodically occurring predetermined tick of a clock. In a distributedTT real-time system, it is assumed that the clocks of all nodes are synchronized toform a global time that is available at every node. Every observation of thecontrolled object is time-stamped with this global time. The granularity of theglobal time must be chosen such that the time order of any two observationsmade anywhere in a distributed TT system can be established from their timestamps with adequate faithfulness [Kop09]. The topics of global time and clocksynchronization will be discussed at length in Chap.

3.Example: The distinction between event-triggered and time-triggered can be explainedby an example of an elevator control system. When you push a call button in the eventtriggered implementation, the event is immediately relayed to the interrupt system ofthe computer in order to start the action of calling the elevator. In a time-triggered system,the button push is stored locally, and periodically, e.g., every second, the computer asksto get the state of all push buttons. The flow of control in a time-triggered system ismanaged by the progression of time, while in an event-triggered system; the flow ofcontrol is determined by the events that happen in the environment or the computersystem.1.6The Real-Time Systems MarketIn a market economy, the cost/performance relation is a decisive parameter for themarket success of any product.

There are only a few scenarios where cost argumentsare not the major concern. The total life-cycle cost of a product can be broken downinto three rough categories: non-recurring development cost, production cost, andoperation and maintenance cost. Depending on the product type, the distributionof the total life-cycle cost over these three cost categories can vary significantly.We will examine this life cycle cost distribution by looking at two importantexamples of real-time systems, embedded real-time systems and plant-automationsystems.181.6.11 The Real-Time EnvironmentEmbedded Real-Time SystemsThe ever-decreasing price/performance ratio of microcontrollers makes it economically attractive to replace conventional mechanical or electronic control systemwithin many products by an embedded real-time computer system.

There arenumerous examples of products with embedded computer systems: cellular phones,engine controllers in cars, heart pacemakers, computer printers, television sets,washing machines, even some electric razors contain a microcontroller with somethousand instructions of software code. Because the external interfaces (particularlythe man–machine interface) of the product often remain unchanged relative to theprevious product generation, it is often not visible from the outside that a real-timecomputer system is controlling the product behavior.Characteristics. An embedded real-time computer system is always part ofa well-specified larger system, which we call an intelligent product.

An intelligentproduct consists of a physical (mechanical) subsystem; the controlling embeddedcomputer, and, most often, a man–machine interface. The ultimate success ofany intelligent product depends on the relevance and quality of service it canprovide to its users. A focus on the genuine user needs is thus of utmostimportance.Embedded systems have a number of distinctive characteristics that influencethe system development process:1.

Mass Production: many embedded systems are designed for a mass market andconsequently for mass production in highly automated assembly plants. Thisimplies that the production cost of a single unit must be as low as possible,i.e., efficient memory and processor utilization are of concern.2. Static Structure: the computer system is embedded in an intelligent product ofgiven functionality and rigid structure.

The known a priori static environmentcan be analyzed at design time to simplify the software, to increase the robustness, and to improve the efficiency of the embedded computer system. In manyembedded systems there is no need for flexible dynamic software mechanisms.These mechanisms increase the resource requirements and lead to an unnecessary complexity of the implementation.3.

Man–Machine Interface: if an embedded system has a man–machine interface,it must be specifically designed for the stated purpose and must be easy tooperate. Ideally, the use of the intelligent product should be self-explanatory,and not require any training or reference to an operating manual.4. Minimization of the Mechanical Subsystem: to reduce the manufacturing costand to increase the reliability of the intelligent product, the complexity of themechanical subsystem is minimized.5. Functionality Determined by Software in Read-Only Memory (ROM): theintegrated software that often resides in ROM determines the functionality ofmany intelligent products.

Since it is not possible to modify the software in aROM after its release, the quality standards for this software are high.1.6 The Real-Time Systems Market196. Maintenance Strategy: many intelligent products are designed to be nonmaintainable, because the partitioning of the product into replaceable units istoo expensive.

If, however, a product is designed to be maintained in the field,the provision of an excellent diagnostic interface and a self-evident maintenancestrategy is of importance.7. Ability to communicate: many intelligent products are required to interconnectwith some larger system or the Internet. Whenever a connection to the Internetis supported, the topic of security is of utmost concern.8. Limited amount of energy: Many mobile embedded devices are powered by abattery. The lifetime of a battery load is a critical parameter for the utility ofa system.A large fraction of the life-cycle cost of many intelligent products is in theproduction, i.e., in the hardware.

The known a priori static configuration ofthe intelligent product can be used to reduce the resource requirements, and thusthe production cost, and also to increase the robustness of the embedded computersystem. Maintenance cost can become significant, particularly if an undetecteddesign fault (software fault) requires a recall of the product, and the replacementof a complete production series.Example: In [Neu96] we find the following laconic one-liner: General Motors recallsalmost 300 K cars for engine software flaw.Future Trends. During the last few years, the variety and number of embeddedcomputer applications have grown to the point that, by now, this segment is by farthe most important one in the computer market.