Modeling and Optimization of Parallel and Distributed Embedded Systems

Chapter 1Introduction^*

The word “embedded” literally means “within,” so embedded systems are information processing systems within (embedded into) other systems. In other words, an embedded system is a system that uses a computer to perform a specific task but are neither used nor perceived as a computer. Essentially, an embedded system is virtually any computing system other than a desktop or a server computer. Embedded systems have links to physical components/systems, which distinguishes them from traditional desktop and server computing [1]. Embedded systems possess a large number of common characteristics such as real-time constraints, dependability, and power/energy efficiency.

Embedded systems can be classified based on functionality as transformational, reactive, or interactive [2]. Transformational embedded systems take input data and transform the data into output data. Reactive embedded systems react continuously to their environment at the speed of the environment, whereas interactive embedded systems react with their environment at their own speed.

Embedded systems can be classified based on orchestration/architecture as single-unit or multi-unit/distributed and/or parallel embedded systems. Single-unit embedded systems refer to embedded systems that possess computational capabilities and interact with the physical world via sensors and actuators, but are fabricated on a single chip and are enclosed in a single package. Multi-unit embedded systems, also referred ...