Edge Computing: A Neural Network Implementation on an IoT Device
Ricardo Barreto, Jorge Lobo, Paulo J. Menezes · 2019
This demonstration showcases the use of reconfigurable logic to implement edge computing for IoT devices able to provide specific information from raw data produced from some sensor, e.g. a camera or microphone, instead of the raw data itself. In what concerns the embedded processing capabilities, the focus is image processing using convolutional neuronal networks (CNN). This approach is clearly distinct from the current trends in IoT devices of using cloud computing to process the collected data. We intend a twist on the established paradigm and pursue an edge computing approach. Since we are targeting small and simple devices, we need some low power solution for the CNN computation. The demonstration will be made on a Terasic DE1 (SoC) reconfigurable system, with a field programmable gate array (FPGA), and hardwired ARM processor to build the IoT device. The collected data from the CNN computation, is transmitted using an IoT protocol to a broker.