Big Data Systems Based IoT Enabling Technologies: Ubiquitous Wireless Communication, Real-Time Analytics, Machine Learning, Deep Learning, Commodity Sensors

Internet of Things

The Internet of Things, or IoT, involves a computing concept or devices that are said to be interrelated with each other. The IoT mainly consists of digital machines, things or devices, and animals or human beings, possessing unique handlers or identifiers (UIDs) without the requirement of human-to-human or human-to- computer interaction. A “thing” in the IoT here either refers to an elderly patient with a pressure monitoring device implanted, a domestic animal possessing sensors, or any object that is allocated with an IP address and is capable of transferring data over a network. Over time, due to the advancement in the areas of applications (such as smart home, healthcare, transportation, vehicle monitoring, and surveillance), certain challenges of IoT enabling technologies have become widespread. Figure 4.1 shows the schematic representation of IoT.

As illustrated in the figure, IoT represents the Internet-able nature of modern physical devices, vehicles, connected devices, or smart devices. Hence, IoT devices possess sensors and software that makes data acquisition and data exchange between users or devices through the Internet possible. Besides, the objects involved in IoT


Figure 4.1 IoT.

are said to be easily controlled in a remote manner with the objective of allowing direct integration with a computer, therefore resulting in economic advantages and user efficiency. Different thoughts have emerged as far as IoT is concerned. Some people think about IoT being connected in terms of computers, tablets, and smartphones. But IoT portrays a world where just about anything is said to be connected and communicated in an intelligent manner. On the contrary, the physical world is predicted to become one of the biggest information systems with the evolution of the IoT. For example, an IoT shopping application easily tracks the location of devices to analyze a person’s shopping preference, habits, and so on. Apart from this, a shopping app would link to a smart fridge that analyzes the past consumption of the user in deciding which food is required and hence sends the grocery list to the person’s smart phone in a smart manner. In other words, a smart fridge automatically orders the products without the interaction or intervention of a human.

On a comprehensive scale, the IoT is found to be applicable to several things, like transportation networks and smart cities, that assist in minimizing waste and enhance the efficiency for several things involving energy usage and also helps in understanding and improving how we work and live. However, the reality in IoT remains in that it ensures endless opportunities and connections. Many of the opportunities can impact or influence today’s environment; hence but the door is also open to many challenges.

With advancement in the field of both wireless technologies and networks, a remote network cannot be perceived without its interrelationship with other networks. Ubiquitous wireless communication (UWC) systems involve different types of wireless heterogeneous networks to achieve intercommunication with any device at any time from any location. The IoT, one the most prominent constituents of UWC necessitates the embedding of computational potential between devices in multiple systems. Besides, an IoT network produces enormous amounts of data in heterogeneous formats, and the applications of big data analytics have increased rapidly in the past few years leading to the next generation of intelligence for big data analytics.

Machine learning proves to be a mandatory tool for big data, as the volume, the variety, and the velocity are frequently improvising the potentialities of typical data analytics. Progress in machine learning is often stimulated through achievements in distinct sub-areas, as it possesses the potentiality to enable intelligent sensors to assimilate data without the need to accurately program the details. Deep learning is a contemporary blooming branch of machine learning that increases the learning performance via multiple layers of processing. This chapter aims to address the concerns about IoT devices, their interrelations, and services they may offer, including an efficient analysis of big data produced by IoT with machine learning techniques and models that will enable IoT to develop and become part and parcel of everyday life. Hence, this chapter first starts with the fundamentals of IoT and then describes the IoT framework and its working. A list of real-time applications of IoT is provided. Certain challenges involved in IoT deployment are also discussed.

Industrial progression of IoT

Figure 4.2 Industrial progression of IoT.

Fundamentals of IoT

In the current era, we are encompassed by plenty of IoT-enabled devices that emit data and communicate through heterogeneous devices uninterruptedly. Moving ahead, in this chapter, the fundamentals required for constructing an IoT application is presented. IoT being a homogenous application employs present-day information technologies, comprising several technical areas, including sensor technology, processing of information, securitizing the information, and communication between users. The four parts of industrial progression of IoT are shown in Figure 4.2.

As shown in the figure, the four parts are identification of the information, information sensing, processing of information, and transmission of the same. Besides, the three elements involved in IoT are sensing of the devices or terminal, network connection, and background evaluation. Among these three, sensing is the basic element. On the other hand, the overall structure of IoT comprises two subsystems: the radio frequency identification (RFID) and information network system. RFID includes a tag and reader and obtains the signal and then sends the signal or packet to the middleware of the information network system via the Internet. The names of objects or things are obtained via Object Naming Service (ONS), whereas the information pertaining to a product is obtained via Electronic Product Code (EPC) interfaces. Therefore, IoT combines several physical product information services based on the setting up of the Internet.

< Prev   CONTENTS   Source   Next >