H. Kopetz and K. H. Kim, "Temporal Uncertainties in Interactions among Real-Time Objects", Proc. SRDS, 1990.  Home/Search   Document Not in Database   Summary   Related Articles   Check

.... protocols is normally far superior: clock driven protocols are, in essence, tight and steady (small and oe) There is a range of applications, in the real time, input output area, where protocols featuring very small and oe are required, to support lock step, time triggered type of operation [13]. Clearly, clock driven protocols are the solution. Alternatively, in application independent distributed systems [3,11,16] running non real time applications, clock less protocols are definitely very well suited, due to the performance advantage. However, clock less protocols also have their ....

H. Kopetz and K. Kim. Temporal uncertainties in interactions among real-time objects. In Ninth Symposium on Reliable Distributed Systems, IEEE, Huntsville, Alabama-USA, October 1990.

....2 Order and synchronism Ordering paradigms assume several forms in a system: total, potential causal, FIFO. Several protocols provide a total order, either real time or not [4, 11, 3, 22] The issue of providing a real time total order is to define a bound on the time needed to secure it [7]. A potential causal order is a much more complex issue, and will be the focus of this paper (the FIFO order is a reduction of the potential causal order to a single sender) The natural order of events in a system is the cause effect relation. Protocols, in order to secure it, tend to order ....

....: oe = max p (T p Dmax Gamma T p Dmin ) 2 These definitions are exemplified in figure 1, where supposedly the runs x and y in figure 1b yield respectively the maximum and minimum delivery times as per definition 4. Steadiness should not be confused with temporal uncertainty, introduced in [7]. Temporal uncertainty is the absolute greatest difference max p;q (T p Dmax Gamma T q Dmin ) In some protocols (e.g. Delta protocols) steadiness is not equal to temporal uncertainty, because they never yield those extremes in the same protocol run or at the same participant. A proof and ....

[Article contains additional citation context not shown here]

Hermann Kopetz and K.H.(Kane) Kim. Temporal Uncertainties in Interactions among Real-time Objects. In Proceedings of the Ninth Symposium on Reliable Distributed Systems, pages 165--174, Huntsville, Alabama, October 1990. IEEE.

....granularity [13] have been pointed out earlier on as disturbance factors of theoretical models based on fully synchronous systems, and perfect and continuous clocks. A model for real time objects and a discussion on the impact of temporal uncertainties on total order protocols has been advanced in [12]. In [22,26] it was proposed to match the granularity of computer derived orderings to the pace at which computer or physical processes evolve (ffi t precedence) A recent work [9] discusses how to match sparse timebases to ffi t precedent sets of events. This paper equates the time and order ....

.... temporal ordering of information; ffl real time and reliability guarantees in communications and processing [10,6,26,23] 4 ffl encapsulating constructs for the distributed and cooperating programs, from the timeliness, dependability and functionality viewpoints, such as real time object groups [12]; ffl system structure for distributed fault tolerance and real time [11,7,25] The sequel of the paper addresses the first two aspects. The following two sections address the roles of time and order. The remaining sections discuss the implications of time and order for distributed applications. ....

[Article contains additional citation context not shown here]

Hermann Kopetz and K.H.(Kane) Kim. Temporal Uncertainties in Interactions among Real-time Objects. In Proceedings of the Ninth Symposium on Reliable Distributed Systems, pages 165--174, Huntsville, Alabama, October 1990. IEEE.

....between nodes, but also atomicity for discrete state variables which are not updated periodically. As an example consider a mode change commanded by an operator. Atomicity may than be required to avoid the case that just a few nodes change mode. Consistency constraints are further elaborated in [20]. We believe that other sets of RT CC s can be defined for applications other than those investigated by us. Hopefully, knowledge about application related RT CC s can be used to improve academic application models. If synchronous execution is not provided, control delays will be time varying ....

Kopetz H., Kim K.H., Temporal Uncertainties in Interactions among Real--Time Objects, Proc. IEEE Computer Society 9th Symp. on Reliable Distributed Systems, Oct. 1990, Huntsville, AL.

....place periodically. We call data elements belonging to this type repetitive . Repetitive data elements carry state information, i.e. they contain real time images of state variables in the environment (for more information on the concept of real time images and objects the reader is referred to [Kop90]) Besides the periodic control application a real time system must be able to cope with events that are outside its sphere of control, e.g. failure events. These events, called chance events, result in a demand for irregular communication. Performance Period The period of change of repetitive ....

H. Kopetz and K. Kim. Temporal Uncertainties in Interactions among Real-Time Objects. In Proc. 9th Symposium on Reliable Distributed Systems, pages 165--174, Huntsville, AL, USA, Oct. 1990. IEEE Computer Society Press.

....within the interaction. We call communication types the possible combinations issued from the number of the tasks participating in an interaction. In addition to the time responsive requirements, we have the fact that their distributed nature makes impossible to have the same global physical time (Kopetz and Kim, 1990). So, this is the cause of the existence of time coherence properties to be verified on the interactions between the distributed cooperative tasks (Thomesse, 1993) So, we need a model that allows us to specify these interactions between tasks and their associated time properties in order to ....

Kopetz, H. and K.H. Kim (1990). Temporal uncertainties in interactions among real-time objects.

....a hard real time system will meet its deadlines. We can only provide a probability that sufficiently few failures will occur, that the fault resilience will remain satisfied, and that the deadlines will thus be met. In actual systems, where faults do occur, there can be no absolute certainties [5]. A very demanding real time application, such as civilian aircraft flight control, may permit a failure rate of 10 Gamma10 per hour. Assuming 10 6 real time deadlines per hour, we can satisfy this stringent system reliability requirement if we can achieve, for each deadline, a probability of ....

H. Kopetz and K. H. Kim, "Temporal uncertainties in interactions among real-time objects," Proceedings of the Ninth Symposium on Reliable Distributed Systems, Huntsville, AL (October 1990), pp. 165-174.

....analysis tools and their interaction with the corresponding design creation tools. Other design evaluation activities and their interactions are described in detail in [Kop91] To describe the activities of a MARS application a transaction oriented model is used. A real time transaction [Kop90b] is a set of correlated actions between a stimulus and a response. It describes the functional behaviour as well as its timing behaviour. A real time transaction can be decomposed into a sequence of real time subtransactions consisting of tasks executions and communication phases between ....

H. Kopetz and K. Kim. Temporal Uncertainties in Interactions among Real-Time Objects. In Proc. 9th Symposium on Reliable Distributed Systems, pages 165--174, Huntsville, AL, USA, Oct. 1990. IEEE Computer Society Press.

No context found.

H. Kopetz and K. H. Kim, "Temporal Uncertainties in Interactions among Real-Time Objects", Proc. SRDS, 1990.

....of Technology [21] 22] and it became clear that a hardware supported fault tolerant clock synchronization is a fundamental building block of a timetriggered architecture. At about that time the important concept of temporal accuracy of real time information was introduced by Kopetz and Kim [23], 24] The TTP C protocol, which includes a clock synchronization service and a membership service, was first published in 1993 [25] A prototype version of the TTA, including a new clock synchronization chip [26] was built in the context of the European PDCS project. This new prototype ....

....and the instant t is smaller than the accuracy interval d acc , which is an application specific parameter associated with the given RT entity. An RT image is thus valid at a given instant if it is an accurate representation of the corresponding RT entity, both in the value and the time domains [23]. While an observation records a fact that remains valid forever (a statement about an RT entity that has been observed at an instant) the validity of an RT image is time dependent and is invalidated by the progression of real time. At the communication network interfaces (CNI) within a node, ....

H. Kopetz and K. H. Kim. Temporal Uncertainties in Interactions among Real-Time Objects. In Proceedings 9th Symposium on Reliable Distributed Systems, pages 165--174, 1990.

No context found.

Hermann Kopetz and K.H.(Kane) Kim. Temporal Uncertainties in Interactions among Real-time Objects. In Proceedings of the Ninth Symposium on Reliable Distributed Systems, pages 165--174, Huntsville, Alabama, October 1990. IEEE.

No context found.

H. Kopetz and K. H. Kim, "Temporal uncertainties in interactions among real-time objects," in Ninth Symposiumon Reliable Distributed Systems, pp. 165--174, October 1990.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC