Creating subsystem interfaces from use case scenarios

There are various ways subsystem interfaces can be created. For example, a common approach is to refine the black box activity diagrams from use case analysis into so-called white box activity diagrams, with swim lanes representing the different subsystems. When control flows cross into other swim lanes, the flow or service invocation is added to the relevant subsystem interface. Another common approach is to do the same thing but use the use case sequence diagrams rather than the activity diagrams. The advantage of these approaches is that they tie back to the use case analysis. It is also possible to create the interfaces de novo by allocating the system features to the subsystems.

This recipe focuses on sequence diagrams. One advantage of this approach is that we can leverage messages on sequence diagrams that have been created by executing the use case models that may not appear on the activity diagram. Furthermore, many...