Embedded Databases for Embedded Real-Time Systems: A Component-Based Approach

In the last years the deployment of embedded real-time systems has increased dramatically. At the same time, the amount of data that needs to be managed by embedded real-time systems is increasing, thus requiring an efficient and structured data management. Hence, database functionality is needed to provide support for storage and manipulation of data in embedded real-time systems. However, a database that can be used in an embedded real-time system must fulfill requirements both from an embedded system and from a real-time system, i.e., at the same time the database needs to be an embedded and a real-time database. The real-time database must handle transactions with temporal constraints, as well as maintain consistency as in a conventional database. The main objectives for an embedded database are low memory usage, i.e., small memory footprint, portability to dif-ferent operating system platforms, efficient resource management, e.g., minimization of the CPU usage, ability to run for long periods of time without administration, and ability to be tailored for different applications. In addition, development costs must be kept as low as possible, with short time-to-market and a reliable software. In this report we survey embedded and real-time database platforms developed in industrial and research environments. This survey represents the state-of-the-art in the area of embedded databases for embedded real-time systems. The survey enables us to identify a gap between embedded systems, real-time systems and database systems, i.e., embedded databases suitable for real-time systems are sparse. Furthermore, it is observed that there is a need for a more generic embedded database that can be tailored, such that the application designer can get an optimized database for a specific type of an application. We consider integration of a modern software engineering technique, component-based software engineering, for developing embedded databases for embedded real-time systems. This merge provides means for building an embedded database platform that can be tailored for different applications, such that it has a small memory footprint, minimum of functionality, and is highly integrated with the embedded real-time system.