Embedded applications are constantly being pushed toward increasing the functionality necessary to achieve autonomy, allowing them to function reliably in all circumstances and under extreme design constraints. Our approach to embedded systems introduces a feedback loop characterizing adaptive systems: the adaptation mechanism monitors system performance and changes the structure accordingly to optimize performance. These self-adaptive systems can be designed and implemented using Model-Integrated Computing. To represent dynamic software architectures, the system is modeled in a generative manner. Here, the components of the architecture are prepared, but their number and connectivity patterns are not fully defined at design time. Instead, a algorithmic description and architectural parameters are provided that specify how the architecture could be generated "on-the-fly". These design-time models are then embedded in the run-time system along with generators that configure/reconfigure the system by changing certain architectural parameters.

This work is sponsored by the DARPA/ITO Active Software Composition program.

For more information, browse the other pages, look at related publications or contact the PI, Akos Ledeczi (615.343.8307).