Model-based System Development : Feature-based System Design Approach : System Partitioning

System Partitioning

The next step in the feature-based design process is the architectural system design. The system is partitioned into subsystems (i.e. ECUs) and then bottom-up the feature/feature interaction models or respective sub-functions are distributed among them. See the System Partitioning, Parsing of Features/Feature Sub-Functions, and Validation of logical Interfaces figure. Although the resulting model is still purely functional, the partitioning is determined almost by implementation considerations, e.g.:

In some cases, architectural design criteria might also be derived from additional Performance Models. These models are based on queuing theory and typically used for throughput analysis and identification of potential communication bottlenecks. Rational Statemate should not be used for this kind of analysis. Rather, you can use dedicated tools such as SES Workbench.

System Partitioning, Parsing of Features/Feature Sub-Functions, and Validation of logical Interfaces
 
Note: At this stage it is not necessary to visualize in detail the communication between the individual ECUs. The Process Buffer is used symbolically as a means for the inter-connection. The broadcasting mechanism is used for external/internal communication.

The subsystem partitioning and feature/function parsing is an iterative process. The particular architectural system design is verified by means of the test scenarios previously used for the verification / validation of the Conceptual Model. For each ECU the respective I/O is recorded.