Editorial for FGCS Special issue on 'Time-critical Applications on Software-defined Infrastructures'

摘要

Performance requirements in many applications can often be modelled as constraints related to time, for example, the span of data processing for disaster early warning [ 1 ], latency in live event broadcasting [2], and jitter during audio/video conferences [3]. These time constraints are often treated either in an "as fast as possible" manner, such as sensitive latencies in high-performance computing or communication tasks, or in a "timeliness" way where tasks have to be finished within a given window in real-time systems, as classified in [4]. To meet the required time constraints, one has to carefully analyse time constraints, engineer and integrate system components, and optimise the scheduling for computing and communication tasks. The development of a time-critical application is thus time-consuming and costly.