Raspberry Pi Based Computing Prototypes: Design, Implementation and Performance Analysis

Nowadays, in Internet of Things (IoT) applications, IoT devices generate and gather data. These data can be handled in different ways. They can be processed by IoT devices directly, by the Cloud, or by the edge servers. In recent years, Raspberry Pi has been developed as a small IoT device for many applications. In this paper, we design and implement various computing models, i.e., Cloud Computing, Edge Computing, and computing on IoT devices, using Raspberry Pis as prototypes. We deploy multiple applications, run hands-on experiments, and analyze the results accordingly. We also compare the performances of these different computing paradigms, and take advantage of the parallel processing capability of Raspberry Pi devices to demonstrate the effectiveness of improving the prototype performance.

I. Introduction

In the Internet of Things (IoT) environment, huge amounts of data have been generated and collected by IoT devices. Traditionally, these data are directly transmitted to the Cloud for processing, because of the limited computing and storage capacities of IoT devices. Nowadays, with more powerful hardware, many IoT devices (e.g., smartphones) can process initial data and send results to the Cloud later.