System on FPGA and Vision
The sub-program System on FPGA and Vision covers research on design of embedded systems. The main focus is on the computational part of an intelligent sensor node.
Field Programmable Gate Array (FPGA) is a computational platform that has a set of resources for logic, arithmetic units, clock synthesis, memories, io, programmable connectivity and even embedded processor cores. This makes it possible to integrate a complete system built on application specific hardware processor units, system busses for communication, embedded soft- or hard core processors, on chip memories, interfaces for external communication and sensor interfaces. This is a very complex platform that requires detailed knowledge in hardware- and software design, integration, testing and verification. These kinds of systems are usually referred to as Systems on FPGA (SoF). The FPGA platform constitutes an attractive compromise of cost, programmability, computing power (GOPS) and power dissipation. In addition, we strongly believe that the integration of a complete system on a programmable device will ease the product maintenance of an intelligent sensor node. A soft IP component will never become obsolete as Commercial On-The-Shelf (COTS) standard components.
The following research activities address questions such as how to best exploit the FPGA architecture for an intelligent sensor node and how to simplify and ease the programming model. The focus is on sensor nodes with high demands on real-time performance, power and energy.
