Reconfigurable Hardware in Fog Computing

Location

Ada, Ohio

Start Date

9-12-2025 1:10 PM

End Date

9-12-2025 1:20 PM

Description

The Internet of Things (IoT) originally promised to make life more convenient by increasing the connectivity and efficiency of systems that were previously only interfaced with direct human connection. With the rise of AI and the increasing cyberphysical world, these devices have not been able to keep up, while often introducing severe performance bottlenecks and security concerns. To address these growing concerns, we developed a testbed to conduct future research on the use of reconfigurable hardware in IoT-style systems. The testbed consists of a Zynq ultrascale FPGA networked to ESP microcontrollers and Raspberry Pis through a standard router. The ESPs and Pis have camera capabilities and act similar to IoT style devices. All networking was done using HTTP for a simple proof of concept. We successfully demonstrated the ability to network FPGAs to IoT-like devices and transfer data between them. Furthermore, we demonstrated the potential of an FPGA to be able to reconfigure itself in runtime to adapt to different conditions. Further research on this topic could realistically use FPGAs to reconfigure themselves to address security concerns as they appear, while operating as a normal hardware accelerator for general use. This research is also influencing the development of labs for the Embedded Hardware Software Codesign course next semester.

This document is currently not available here.

Share

COinS
 
Dec 9th, 1:10 PM Dec 9th, 1:20 PM

Reconfigurable Hardware in Fog Computing

Ada, Ohio

The Internet of Things (IoT) originally promised to make life more convenient by increasing the connectivity and efficiency of systems that were previously only interfaced with direct human connection. With the rise of AI and the increasing cyberphysical world, these devices have not been able to keep up, while often introducing severe performance bottlenecks and security concerns. To address these growing concerns, we developed a testbed to conduct future research on the use of reconfigurable hardware in IoT-style systems. The testbed consists of a Zynq ultrascale FPGA networked to ESP microcontrollers and Raspberry Pis through a standard router. The ESPs and Pis have camera capabilities and act similar to IoT style devices. All networking was done using HTTP for a simple proof of concept. We successfully demonstrated the ability to network FPGAs to IoT-like devices and transfer data between them. Furthermore, we demonstrated the potential of an FPGA to be able to reconfigure itself in runtime to adapt to different conditions. Further research on this topic could realistically use FPGAs to reconfigure themselves to address security concerns as they appear, while operating as a normal hardware accelerator for general use. This research is also influencing the development of labs for the Embedded Hardware Software Codesign course next semester.