Mixed-criticality (MC) systems are of particular interest today. The adaptive scheduling of tasks on such systems attracts a lot of attention because of the complexity of the activity due to the high number of parameters and constraints and the need to find solutions as the scale of the problem increases.

In this long paper, the authors propose an adaptive MC scheduling analysis and optimization able to fulfill the designers’ constraints and optimize the adopted objective function, while keeping the complexity of the design space exploration under control. The adopted solution trades off accuracy (in a moderate way) in order to manage problems that are almost twice the size of the problems that can be handled by existing solutions. To do so, the authors formally introduce the elements of the problem and demonstrate, through a series of theorems, how it is possible to work on a smaller problem.

The experimental validation is presented with an in-depth analysis of the quality of the proposed approach against the reference work performed by Baruah et al. in 2011 [1], used as a baseline for both accuracy and scalability; results well support the authors’ claims.

The problem is a complex one, and its formulation and solution require a formal approach that makes the paper particularly “heavy” to read, consisting mainly of theorems, lemmas, and proofs. The application to two case studies allows the reader to qualitatively and quantitatively appreciate, at the end of the long formalization, the presented work.