ABSTRACT
Context management is one of the major issues that is identified and discussed in the industry as well as in academia. The connectivity context is handled inadequately by many mobile applications because they cannot switch flexibly between offline and online situations. Mobile application developers are often unaware of suitable architectural designs for their types of mobile application and of how transactions and data can be managed across changing connectivity states. However, users expect their mobile applications to run reliably even when the connection is limited or interrupted intermittently.
A promising solution to meet this requirement is the model-driven design of online- and offline-capable (i.e., connectivity-aware) mobile applications because it separates the technical concerns from the business concerns. Here, we present a design method and a tool environment that support mobile application developers in creating online- and offline-capable mobile applications on a high level of abstraction. An application model (app model) that captures the underlying data structure, behavior, and graphical user interface (GUI) of the mobile application is used to create online- and offline-capable mobile applications. Moreover, this model-driven approach provides app model analysis and simulation to predict the number of successfully executed transactions—known as throughput—for different connectivity levels and numbers of mobile application users.
