The IoT or the internet of things started as a technology to connect everyday objects over the internet, which has evolved into something big and invaded into every single aspect of our lives. As technology is gaining momentum, IoT-based smart devices usage among users is expanding, which generates massive data at our disposal across various domains. The authors have systematically studied the taxonomy of data analytics and the benefits of using advanced machine learning techniques in converting data into valuable assets. In the studies, they have identified and did due diligence on different smart home systems, their features, and configuration. During this course of study, they have also identified the vulnerability of such a system and threats associated with these vulnerabilities in a secure smart home environment.
Top1 Introduction
Recent advancements in the internet of things and improvements in computing devices have made communication between devices easy. It is fast becoming a topic of social, technical, and economic significance. A wide range of consumer goods, mobile devices, cars, industrial components, utility items, sensors, and other objects are combined with internet connectivity and powerful data analytics capabilities to transform the life around us. The forecast says that by 2030 total connections to the internet will be more than 50 billion and globally impacting more than $11 trillion. McKinsey stated a 300 percent rise in linked IoT devices in 2013 annual report and valued the future economic effect of IoT at $2.7 trillion to $6.2 trillion annually by 2025 (Manyika et al., 2013; Manyika et al., 2015). Because users are now very dependent on different applications over the internet, technologies like the Internet of Things (IoT) will be more widespread in coming years. The embedded technology that includes wired and wireless communication along with sensors, actuators, and physical devices is known as IoT. The aim of smart computing devices is augmentation and effortlessness in the experience provided to the users. Hence, IoT is fast becoming a key source of new data, that demands storage and analysis needs. The IoT offers endless ways to connect everyday objects which a common person uses inside and outside the home. In terms of innovation, this area is wide open accelerating demands for machine learning and other interconnecting technologies. It may be an encouraging time for creative people, partially because understanding these interconnections is still arduous. The system offers both prospects and potential security problems.
Smart devices like smart phone, vehicles, temperature control systems, smart elevators, healthcare devices and automation systems are making our life easier and better. Large number of IoT devices being used and their support systems generates massive amount of data that needs storage and processing at cloud storage centers. The massive data generated in the process cannot used for knowledge discovery unless processed by advanced machine learning techniques that can handle it properly. The IoT application domain range from social media, smart healthcare, smart e-agriculture, smart electricity, and smart vehicles. Moreover, the IoT success depends on specific protocols to interconnect such as application-layer protocols that interacts with users directly. The protocols such as Constrained Application Protocol (CoAP), Hyper Text Transfer Protocol (HTTP), Message Queuing Telemetry Transfer (MQTT), and Advanced Message Queuing Protocol (AMQP) (Karagiannis et al., 2015; Zschörnig et al., 2020) are responsible for communication between devices.
Figure 1. Growth of IoT market (in billion dollars) by statista.com (Forecast end-user spending on IoT solutions worldwide from 2017 to 2025, n.d.) accessed on 30 December’2020
Intelligent living occurs, as smart devices, apps and utilities work together to create an environment that surrounds us. The IoT based devices help people live their lives comfortably and securely, so that they can concentrate on what really matters. The Statista (Forecast end-user spending on IoT solutions worldwide from 2017 to 2025, n.d.) forecast predicts that global demand for smart devices is growing, with shipments to surpass three billion by 2023 for devices such as smartphones, PCs, laptops, wearable bands, smart speakers, and smart personal audio systems. The demand for smart speakers is expected to grow with shipments set to exceed 39 million units in 2023, because speakers can serve as command centers in smart homes.
The contribution of the chapter are as follows:
- •
review different machine learning application for IoT data.
- •
Taxonomy of machine learning algorithms.
- •
The features of IoT data in real-time.
- •
Smart city as the application of IoT.
- •
Challenges in secure smart home systems and IoT data analytics.