Development Cost Optimization for Multi-Functional Mixed-Criticality Embedded Systems

Abstract

In some industrial embedded systems, functions with different safety integrity levels (SILs) are integrated into the same hardware platform to share resources, and usually, the schedulability of applications should be met first. Meanwhile, development costs' optimization of applications is also an important requirement, because some industries, such as automotive, are cost-sensitive for the mass market. To minimize the development costs of applications while satisfying their schedulability, an optimization method is presented based on a key path tabu search algorithm combined with a message movement strategy. The key path tabu search algorithm uses the tree-based tabu list to find a schedulable solution with low development cost. The message movement strategy is proposed to significantly reduce the development cost caused by the recursive elevation of the SILs of tasks for the sharing of partitions. The results of experiments using real-life benchmark and simulated benchmark show that the proposed method is effective.