ACM Computing Surveys 28A(4), December 1996, http://www.acm.org/surveys/1996/RamamrithamPredictability/. Copyright © 1996 by the Association for Computing Machinery, Inc. See the permissions statement below.

Krithi Ramamritham

1. Real-Time systems come in many shapes and forms and are deployed in many different situations, each having a different set of needs.
2. Requirements imposed on different parts/layers of a system can vary drastically.

Perhaps because of these reasons, there continues to be a big gap between the types of abstract systems that academics have tended to study and real-world real-time systems. The latter are able to survive and continue to operate even when not all timing constraints are met. This is especially true for applications with less stringent ``Quality of Service'' requirements, such as multimedia. Early real-time literature had categorized such systems as soft-real-time systems but they are not quite the same and hence many interesting issues have gone unexplored:

1. Precise specification of the predictability requirements. Most are likely to be probabilistic in nature.
2. Developing scheduling and resource management schemes for (sub)systems with demonstrable predictability properties.
3. Given subsystems with known predictability properties, developing schemes to compose them into larger systems such that the predictability properties are also composable. This is essential to make real-time system components reusable.

Addressing these challenges demands elevating predictability to the level of a first-class requirement in the design and development of each component of a real-time system and developing design and analysis tools to satisfy the predictability properties as well as to understand the trade-offs, for example, between predictability and resource utilization.

Permission to make digital or hard copies of part or all of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. Copyrights for components of this work owned by others than ACM must be honored. Abstracting with credit is permitted. To copy otherwise, to republish, to post on servers, or to redistribute to lists, requires prior specific permission and/or a fee. Request permissions from Publications Dept, ACM Inc., fax +1 (212) 869-0481, or permissions@acm.org.