| G. Fohler. Analyzing a Pre Run-Time Scheduling Algorithm and Precedence Graphs. Research Report 13/92, Institut fur Technische Informatik, Technische Universitat Wien, Vienna, Austria, September 1992. |
....select the best then. For this reason bus access planning is performed by heuristic search [Ree93] Much work has been done on deterministic scheduling of task sets on multiprocessor systems. Most of the papers also incorporate static bus access planning as side effect of task scheduling, e.g. [Ram90, Foh94, Foh92, Ver95a, Ver95b]. Yet only little attention has been paid to independent, explicit planning of bus access. An approach is presented in [Tin94] The authors of this paper focus on guaranteed latencies for critical messages in a CAN based system. In this paper we present the requirements for a time triggered ....
G. Fohler. Analyzing a Pre Run-Time Scheduling Algorithm and Precedence Graphs. Research Report 13/92, Institut fur Technische Informatik, Technische Universitat Wien, Vienna, Austria, September 1992.
No context found.
G. Fohler. Analyzing a Pre Run-Time Scheduling Algorithm and Precedence Graphs. Research Report 13/92, Institut fur Technische Informatik, Technische Universitat Wien, Vienna, Austria, September 1992.
....the very minor modification to the run time dispatcher to check for black out times. We have implemented a simulator for the run time execution of mode changes. 5 Constructing Schedules We present an approach to off line scheduling of mode changes on the basis of the MARS pre run time scheduler [5, 12]. We start by summarizing briefly the base algorithm and go on to describe the modifications required in order to schedule the mode changes. 5.1 Single Mode Schedule Construction We view scheduling as a kind of problem solving: The precedence graphs, the number of CPU slots consumed by each ....
....the domain of scheduling, however, we do not need an optimal solution, i.e. the schedule with the shortest response time, but any one feasible will do. Thus we have implemented some modifications to IDA in order to reduce run time (see figure 5) More details about the implementation are given in [12]. 7 A pseudo code listing of the implementation is given is figure 5. 10 ITERATION( if (STARTDEPTH = 1) threshold = MAX; else threshold = STARTDETPH; STARTDEPTH is a user defined threshold for the first iteration while(DEEPEN (startnode) not done) threshold = threshold ....
G. Fohler, "Analyzing a pre run-time scheduling algorithm and precedence graphs," Research Report 13/92, Institut fur Technische Informatik, Technische Universitat Wien, Vienna, Austria, September 1992.
....it is accepted. 6 Analysis We have implemented the described mechanisms and have run simulations for various scenarios. The system simulated consists of 4 processing nodes, connected via a TDMA driven communication medium. Some 1600 task sets were generated and scheduled by the MARS scheduler [7]. We randomly generated aperiodic task loads for each of these, so that the combined load of static and aperiodic tasks was set to 20 , 40 , 60 , 80 , and 100 . Aperiodic tasks were simple and preassigned to processing nodes. The deadline for the aperiodic tasks was set to their MAXT and two times ....
G. Fohler. Analyzing a pre run-time scheduling algorithm and precedence graphs. Research Report 13/92, Institut fur Technische Informatik, Technische Universitat Wien, Vienna, Austria, September 1992.
....execution of a set of periodic activities that have the task characteristics described above. Many scheduling algorithms have been proposed for statically scheduling precedence constrained tasks in distributed systems, where task deadlines are equal to or less than their periods (for example, [18, 14, 6]) Given a set of periodic tasks, such a scheduling algorithm attempts to construct a schedule of length lcm, the least common multiple of the task periods. The schedule specifies the exact times at which the tasks will begin execution 1 . A real time system with the given set of tasks then ....
....repeating schedule is equal to the lcm of the periodic tasks. 2. 1 Scheduling Tasks with Deadlines not Greater than Periods Let us have a closer look at how static schedules are constructed, that is, how the shortest repeating schedule is determined with static scheduling algorithms, e.g. [18, 14, 6]. Given tasks with deadlines, precedence constraints, resource requirements (such as specific requirements for processors, memory, etc. and replication constraints, these algorithms determine a static schedule which specifies the exact times at which the tasks will begin execution. In contrast, ....
G. Fohler. "Analyzing a pre run-time scheduling algorithm and precedence graphs". Research Report 13/92, Institut fur Technische Informatik, Technische Universitat Wien, Vienna, Austria, September 1992.
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