Real‐Time Resource Management Techniques

Abstract

Abstract Real‐time systems are used to monitor and control complex physical processes ranging from the civilian problems of aircraft traffic control, manufacturing and space flight, to the military applications that arise with modern warfare systems. Unlike general purpose computer systems, where resources must be managed to provide adequate system responsiveness for all tasks on the average, the underlying physical processes being controlled by a real‐time system impose explicit timing requirements on the tasks processed by the computer system. These timing requirements are an integral part of the correctness and safety conditions of a real‐time system. Although real‐time systems must manage many resources including memory, I/O, files, etc., it is the management of time which distinguishes real‐time systems from ordinary computer systems. In this article, we address resource management for real‐time uniprocessor‐based systems and focus exclusively on time management as applied to processor scheduling. Proper resource management, to ensure that tasks meet their timing requirements, is a vital aspect of a real‐time system design where failure to meet a task timing requirement causes a timing fault to occur. speed alone is not enough. Proper resource management techniques must also be used, A guiding concern in real‐time system resource management is predictability , the ability to determine for a given set of tasks whether the system will be able to meet all of the timing requirements of those tasks. Real‐time systems must be highly fault tolerant and behave properly in the presence of timing faults or transient system overloads. Embedded real‐time systems must schedule diverse activities to meet the timing requirement imposed by the physical environment. While this issue may be a difficult problem, the real‐time system developer works in a tightly constrained environment and, typically, has the advantage of knowing the entire set of tasks that are to be processed by the system and the conditions under which each might be run. An algorithmic‐based scheduling methodolgy allows the developer to combine task‐set characterization data with the associated timing requirements, to determine through the use of schedulability criteria whether or not the task set can be scheduled to meet its timing requirements. Consequently, a scheduling theory can make the timing properties of the system predictable. This methodology allows the determination, analytically, as to whether the timing requirements of a task set will be met, and if not, which task‐timing requirements will fail. An iterative process can then be used to modify certain tasks to achieve the schedulability of the entire task set.