Real-time systems (including model-based, scheduling, optimal, heuristics)

Many safety-critical application subsystems demand hard real-time requirements, where the achievement of stability and safety depends on the completion of activities (e.g., reading of sensor values, performing computations, message transmission) in the defined time.

In such hard real-time systems, missed deadlines represent system failures with the potential of consequences as serious as in the case of providing incorrect results. Thus, hard real-time systems must ensure a guaranteed response even in the case of peak load and fault scenarios. A timing and resource analysis must assess the worst-case behavior of the system in terms of communication delays, computational delays, jitter, end-to-end delays, and temporal interference between different activities.

Scheduling algorithms serve for the temporal and spatial allocation of resources, while meeting deadlines and satisfying functional (e.g., precedence constraints) and extra functional constraints (e.g., reliability, energy).

Our expertise encompass:

  • Real-time scheduling algorithms (e.g., MILP, genetic algorithms, list scheduling)

  • Multi-criteria optimization (e.g., timing, energy, reliability)

  • Modelling of platforms and applications

  • Timing analysis

  • Meta-scheduling and mode changes