Analytical review of existing techniques IoT based Home Automation and securing its data over Edge Computing using Machine Learning

Abstract

Future home designs heavily rely on the smart home concept. The primary goal of Home automation is to provide pleasant, secure, efficient, and easy management or monitoring of home appliances, using sensors that save energy too. For controlling the home appliances, existing systems use Bluetooth, Zigbee, GSM Module, WLAN, or Internet Technology. Home automation systems have been further enhanced using Arduino, Node MCU, Raspberry Pi, and WIFI Module. When billions of devices are online and this data-rich world enabled by the Internet of things is fully operational, cloud computing won't be able to handle the workload. Every IoT sensor produces a significant amount of data. Data is promptly transferred to a central cloud database in cloud computing, where it is managed and stored. But occasionally, a response may be interrupted or delayed owing to a network error or a poor internet connection simply because the data centre is too far away from the device. This problem is solved by the use of edge computing, IoT device itself or the nearby node network is behind the data processing and can response in an exact manner if a specific action is needed. In the proposed paper a detailed review conducted regarding existing technologies and existing home automation system (HAS). With the same reference to IoT, new architecture needs to be developed and providing data security using the edge computing instead of Cloud Computing.