Define test context

The first step in performing Component Integration test is to define what will be tested, the System Under Test (SUT). Since the goal of component integration testing is to confirm that all components work together correctly in the target environment, the first task is to confirm that the software components integrate correctly. Perform these tests on the desktop using unit testing principles. The reason for this approach is simple: it reduces the number of sources of defects to the minimum set. Once the software-only components have been tested, move to testing the interactions with hardware and testing the components on the target hardware.

Often, component integration testing requires that the tests run on the target device or a simulator. This environment restriction requires a special software build with all required components. For example, testing the interactions of a component with hardware might require the hardware, drivers, signal conditioning, and the component. If all required parts are not available, it may be possible to use a simulator or build special testing components.

Define test cases

Identify test cases by identifying the functionality you are building. Each test case defines a specific capability. This functionality may be described by a use case, an activity diagram or flow chart, a state chart, or a set of requirements. For each test case, define one or more test scripts so that the capability is fully exercised. It is important to specify the scenario completely, including exception handling conditions.