From Kernel to Cloud: A Concise Comparative Study of Practical IoT Operating Systems

The growth of the Internet of Things (IoT) is expected to rapidly interconnect tens of billions of new devices, many of which will be connected to the Internet. These IoT devices span a wide range, from high-end devices that can utilize traditional Operating Systems (OSs) to low-end devices with stringent resource constraints, i.e., memory, processing power, and power supply. Effectively developing, deploying, and maintaining large-scale IoT systems requires appropriate OSs that can cater to the diverse needs of these resource-constrained IoT devices. Due to the significant role of OSs on the performance of emerging IoT applications, this study provides a detailed comparative study of the top practical IoT OSs over the past few years. We investigate various OSs from the unique perspective of their suitability for embedded devices and edge nodes in IoT applications, providing a comprehensive understanding of how their architectural features and design choices align with the requirements of resource-constrained IoT environments. Furthermore, by conducting a comprehensive set of real-world implementations, this comparative study is accompanied by a deep evaluation of IoT-based and non-IoT-based versions of the most widely-used OS, in terms of power consumption, CPU utilization, and memory usage. These evaluations give valuable insight into how various aspects of the OS design could affect its performance to be selected as the underlying embedded or edge devices OS.