Connected systems with edge, fog and cloud networks bring many uncertainties
The demand for new and more complex applications is constantly growing. By moving individual systems to the cloud or adjacent edge nodes, the overall system becomes more flexible, resource-intense functions are easier to implement and life cycle management improves.
With higher level systems in which many individual systems interact, the architecture must take into account the fluctuating quality of the connections and react accordingly. Especially in dynamic environments where new participants such as devices, machines, applications or networks are constantly added or removed, permanent adaptation is imperative.
During the development phase, the ecosystem consisting of end-user devices (smartphones, connected vehicles…) and edge/fog/cloud networks can be used as a test system only with extensive effort. On the other hand, testing in a real environment is time and cost consuming. Furthermore, edge clouds have not been available for commercial use to date.
Despite the many uncertainties, software architectures must ensure that the complex overall system functions in a dependable manner. Especially when it comes to safety-critical functions, strict requirements must be adhered to, such as the real-time capability, availability and dependability of the connected functions.