Design Abstraction for Autonomous Adaptive Hardware Systems on FPGAs

Adaptive hardware is gaining importance with the emergence of more autonomous systems that must process large volumes of sensor data and react within tight deadlines. To support such computation within the constraints of embedded deployments, a blend of high throughput hardware processing and adaptive control is required. FPGAs offer an ideal platform for implementing such systems by virtue of their hardware flexibility and sensor interfacing capabilities. FPGA SoCs are specifically well suited offering capable embedded processors that are tightly coupled with a flexible high performance FPGA fabric. This paper explores existing work on adaptive hardware systems before proposing a general model and implementation approach tailored towards these modern FPGA architectures, concluding with pointers for research in this emerging field.