FPGA-based Embedded Sensor Architecture for High-Speed Data Acquisition Systems
Abstract
High-speed data acquisition systems play a vital role in many arenas from hospital-based sensor systems to factory automation. Such systems have requirements of high throughput, low latency, and real-time performance capabilities, and a high-performance processing architecture is required. Field Programmable Gate Arrays (FPGAs) offer a powerful front-end competitive alternative for high-speed data acquisition because of the parallel processing inherit to the architecture, recent developments in ease of reconfiguration, and the ability to interface with runtime processing. This paper presents embedded sensor architecture based on FPGA technology has invented to assist with high-speed data acquisition applications. The architecture provides a framework to integrate multiple sensors, signal conditioning, and analog to digital converters (ADCs) on a single, reconfigurable FPGA platform. The system has performance characteristics of high throughput, extreme low-latency, and scalable processing. By modularizing the features, the system can be easily adapted to different applications. Evaluation clearly showed that the system was able to consistently outperform misleading timing results from conventional data acquisition methods using microcontrollers or DSPs. This illustrates a possible architecture to enable sensor-level interfacing with high-speed data analytics in one device, and it serves as an example of a better integrated sensor and real-time data handling system to address fast-changing demands with a flexible future proof design. The proposed architecture also has additional possibilities as a system to support time-critical applications such as telemetry, automotive radar event scanning, or diagnostics.