Programming Embedded Computing Systems Using Static Embedded SQL

The information technology boom in the last decade has made embedded computing systems increasingly common. This has fueled the need for increased automation for large parts of the software development process of such systems. However, such automation must account for the fact that embedded software may require guarantees on response times and can have limited memory available for storing code and data. In this paper, we show how parts of the software can be written in a declarative programming language such as SQL. This is challenging because (1) SQL is a declarative language that abstracts any consideration of execution time, (2) most commercial SQL engines have a large footprint that cannot be stored on an embedded device, and (3) most SQL operations can be executed in satisfactory time only when potentially large amounts of additional storage is available for auxiliary structures, such as indices and materialized views. The paper shows how these challenges can be addressed by using the following strategies.