Quantifying the impact of uncertainty in embedded systems mapping for NoC based architectures

We describe a modeling framework to capture and account for uncertainty in design parameters embedded systems. then develop an uncertainty-aware solution the problem of mapping systems that uses Network-on-Chip (NoC) based architecture platforms. The is formulated as multi-objective - reliability, performance, energy consumption optimization problem. To solve this problem, we propose on NSGA-II genetic algorithm Monte Carlo simulation techniques. implemented computer-aid tool can generate robust 3D Pareto frontiers space formed by objectives consumption. Comparison several state-of-the-art models solutions indicate significant differences actual values attribute interest exist when one considers parameters. For example, case with reliability only objective, 10% lead 10.06% difference MTTF estimation. In execution time objectives, 2D due be up 7.9%. These are important because they mislead overall process toward suboptimal points. DESUU-NOC implements proposed has main feature contribution paper ability comprised