Software organization to facilitate dynamic processor scheduling

On scheduling garbage collector in dynamic real-time systems with statistical timing assurances

“…We estimate the statistical properties-e.g., distribution, mean, variance, of job execution time demand rather than the worst-case demand because: (1) applications of interest to us (Clark et al 2004;Clark et al 1999) exhibit a large variation in their actual workload. Thus, the statistical estimation of the demand is much more stable and hence more predictable than the actual workload; (2) worst-case workload is usually a very conservative prediction of the actual workload (Aydin et al 2001), resulting in resource over-supply; and (3) allocating execution times based on the statistical estimation of tasks' demands can provide statistical timing assurances, which is sufficient for our motivating applications.…”

“…Another important distinguishing feature of most of these systems (at least the ones of primary interest to us) is their relatively long activity execution time magnitudes, compared to those of conventional hard real-time subsystems-e.g., in the order of milliseconds to minutes. Some examples of such dynamic systems that motivate our work include (Clark et al 1999(Clark et al , 2004.…”

Garbage Collector Scheduling in Dynamic, Multiprocessor Real-Time Systems

“…Some examples of such systems that motivate our work include [2], [3], [4]. For example, in [2], Clark et al discuss an AWACS tracker application which collects radar sensor reports, identifies airborne objects (or "track objects") in them, and associates those objects to track states that are maintained in a track database.…”

“…Here, each instance of a track association activity must complete as soon as possible, but before a certain time. In [3], Clark et al discuss the Mission Data System (MDS) of the NASA/JPL Mars Science Lab Rover robot application with similar timeliness semantics. Additional key features of this application include processor cycle overloads and activity time scales (e.g., frequency of constructing MDS schedules) that are of the order of minutes (AWACS [2] also has these features).…”

Utility Accrual Real-Time Scheduling under Variable Cost Functions

“…Embedded real-time systems that are emerging in many domains such as robotic systems in the space domain (e.g., NASA/JPL's Mars Rover [4]) and control systems in the defense domain (e.g., airborne trackers [2]) are fundamentally distinguished by the fact that they operate in environments with dynamically uncertain properties. These uncertainties include transient and sustained resource overloads due to context-dependent activity execution times and arbitrary activity arrival patterns.…”