Adaptive Communications Scheduling for Embedded Systems

Model-based technologies have been traditionally applied to embedded systems design from a functional perspective. Real-time scheduling issues and the associated modeling concerns are becoming increasingly more prevalent. Traditional schedulability analysis is being applied to modern model-based systems to generate schedulable systems. These schedulability analysis techniques are also applied to communications (e.g., network) scheduling. With adaptive communication scheduling, analysis need not simply decide the “pass/fail ” criteria of a particular problem, but can explore the space of a variety of solutions. This paper discusses adaptive communications scheduling for embedded systems. A fuzzy logic based scheduler is used in the system to allow for flexibility in experimentation of different scheduling algorithms. Various examples are presented that show how a model-based adaptive scheduler can be used to achieve a system that adapts to environment conditions at system design, synthesis, and runtime. 1