A reliability indication method for constant-time admission control test

Admission control test aims at determining whether an arriving task can be scheduled together with the tasks already running in a system, so that it can prevent system from overload and collapse. Although a number of schedulability tests have been studied, they are incomparable if the determination conditions are different. In order to keep high system utilization, the problem of choosing a reliable test attracts our attention. In this paper, we introduce a concept of test reliability to evaluate the probability that a random task set can pass a given test, and define a reliability indicator to show the test reliability. The proposed method is applied in several existing admission control tests. Results show that this method can be used as a criterion to compare the performance of different tests. The most applicable condition can be selected among several alternatives using the reliability indicator, if system parameters are determined in advance.