Embedded system for the communication and monitoring of an electric microgrid using IoT

This work presents a system for the communication and monitoring of an electric microgrid. The system presents a novel architecture using a wireless sensors network with a sensor node based on a Digital Signal Controller, a base node based on an Embedded System and a Human Machine Interface implemented on a mobile application. These controllers are composed with a processing unit capable of executing digital signal processing algorithms and a 16-bit microcontroller unit with several peripherals. The sensor nodes are configured with a sampling frequency of 5120 Hz, each sample is 12 bits in size therefore 10240 bytes are transferred per second. The sensor nodes are using a WiFi communication module with an embedded 32-bit processor where the TCP/IP stack resides. The WiFi module is configured using AT commands via UART Interface at a baud rate of 115200. The base node is based on an ARM Cortex A53 four-core processor with a custom minimal Linux distribution developed using the Yocto Project. In the base node, two daemon servers are booted up using System V, in addition, both are using a client-server architecture using TCP sockets. One of the servers obtains the data from the sensor nodes and stores it, while the other is tasked with handling mobile devices requests to visualize the stored data. The complete system is designed to work autonomously according to the IoT concept. The experimental results also demonstrate that the embedded servers receive all samples without loss, from sensor nodes, when a block length of 2, 4 and 8 bytes is selected. Therefore, the proposed architecture can perform real-time monitoring in a microgrid with embedded systems of low cost and low power.