Big Data Analytics

Big data analytics investigates huge amounts of data to extract the patterns that are considered to be hidden, identify the correlations between the hidden patterns, and so on. With the advent of and improvement in technology, data are said to be analyzed in an effective manner via big data analytics. In other words, it helps organizations and establishments to use the valuable data and acquire new opportunities with the identified data. Hence, a smarter business with higher profit rate is said to be achieved by making both the seller and buyer happy. In this chapter, we first define data science and big data. Then, the technologies involved in big data are described followed by an explanation on the applications of big data in several domains. Finally, the opportunities and issues related to big data analytics are discussed.

4.2.1 Data Science and Big Data

Data science deals with both structured and unstructured data. In other words, it is referred to as the field that deals with everything related to data. It starts from data cleansing and proceeds to data preparation and finally data analysis. Figure 4.4 shows the three crucial regions of data science.

As illustrated in this figure, data science [2] is simply referred to as the integration of statistics, mathematics, and logical programming in situations involving problem-solving; capturing data; and the potentiality to view things in a different manner. However, big data refers to huge data volumes that are not said to be processed in an effective manner using the conventional form of processing that exists in the market.

4.2.2 Technologies Involved in Big Data

Big data analytics [3] involves the combination of different materials and methods in processing. The technology is made effective by the collective use of these materials and methods by organizations to gain relevant results for strategic management and implementation in a timely and precise manner. Several organizations consider different technologies involving big data for their success, which heavily depends on the organizations using it and the type of service they provide. However, the nine most prevailing technologies involved in big data used by competing organizations are

Structure of data science

Figure 4.4 Structure of data science.

■ Data pre-processing

■ Data virtualization

■ Data integration

■ Data quality management

■ In-memory storage

■ Predictive analysis

■ Knowledge discovery tools

■ Stream analytics

■ Distributed storage.

4.2.3 Applications of Big Data

As per the strategy followed in business market, industries missing the opportunities related to big data are said to miss innovation, competition, and productivity. In other words, the tools and technologies involved in big data [4] aid in assisting industries to transform huge amounts of data in a swift manner, therefore assisting them to improve production efficiency. Hence, the applications related to big data are found to create a new era in all aspects of life and in all sectors. Figure 4.5 shows examples of big data applications in different industries.

As illustrated in the figure, the applications of big data range from banking, to insurance, to education and media, to healthcare and transportation. Hence, the big data is found to be of profound significance and is widely applied in various sectors.

4.2.4 Opportunities and Issues in Big Data Analytics

With the evolution of human civilization, data collection and perceiving critical information are the two foremost things to be considered. From prehistoric data storage to the present-day sophisticated technologies of Hadoop and MapReduce, different methods and materials have been used for storing and analyzing data. However, several opportunities and issues in big data analytics have to be handled in a proper manner for productive results. Advantages of Big Data Analytics

Some of the advantages of big data analytics are

■ Lower cost

■ Newer business opportunities

■ Application in all sectors of organizations

■ Efficient decision-making in a competitive market

■ Applicability to both unstructured and semi-structured data. Issues in Big Data Analytics

Some of the issues of concern in big data analytics are as follows:

■ Hadoop is harder to understand.

■ Rise of unstructured and semi-structured data.

■ Dealing with scalability.

■ Validating and securing big data.

■ As big data are said to be created in a swift manner, organizations also need to respond to them in a swift manner.

Wireless Communication and IoT

One of the crucial segments of the IoT framework is a wireless communication system. The wireless communication system acts as a platform for two-way communication between the users, specifically for collecting data or information and supervising delivery of messages in a continuous manner with less human intervention. Hence, it is said to be applied to different IoT applications. Some of the IoT applications with wireless communication system are essential industries, such as power grids and oil fields, and chores involved in routine life in smart cities. This chapter introduces wireless communication and IoT. In this chapter, first the design and architecture of wireless communication are described. We then discuss modern ubiquitous wireless communication with IoT design structure. Finally, the implementation of communication models for IoT is discussed.

4.3.1 Introduction

IoT is a new movement. With the objective of achieving a ubiquitous computing system, several materials and methods work in an efficient manner in the IoT. Some of them are RFID, WSN, and cloud computing (CC). Among all the technologies and methods, these three - RFID, WSN, and CC — play a prospective role in the IoT. With the inception of IoT, WSN is considered to be its most salient area. The main goal of IoT is to construct a worldwide network with the aid of all the probable objects. Besides, WSN is considered to be the most truly aiding technology that lets a user understand and achieve the real objective of IoT. The main idea behind wireless communication [5] and IoT is to associate the sensing layer and network layer in the IoT.

4.3.2 Architecture of Wireless Communication

Figure 4.6 shows the architecture of wireless communication. As illustrated in the figure, two different types of nodes are present, and they are in two different colors for differentiation, with blue-colored nodes representing the normal nodes and violet-colored nodes representing the nodes being detected as the sink nodes SN for performing a specific task. With the sensor being detected, nodes in the neighboring area are said to be connected with each other in a wireless mode for better communication.

Finally, the sink node obtains the data or the data packet from the closest neighboring node. This therefore forms the basis of a self-sufficient network in WSN.

4.3.3 Modern Ubiquitous Wireless Communication with IoT

The ubiquitous wireless communication is also referred to as pervasive communication. It is the dissemination of communications framework and wireless technologies throughout the environment to enable ceaseless connectivity. Amongst them, one of the most distinguished elements of pervasive communication is the IoT. It involves the embedding of computational potentialities into everyday objects throughout the environment, with which they are connected to each other. Ubiquitous networking refers to the integration of both wired and wireless technologies that underpin communication between several objects involved. On the other hand, ubiquitous networks in home/industry environments are extending the model of connectivity to all types of objects or things that are found to be heavily in use to construct both home and industrial network services. This type of technology in both home and industry uses either thing-to-thing or thing-to-gateway connections for accessing the IoT. In this way, the data are said to be stored for future use and for accessing varied services, such as remote home sensors, remote industry sensors, etc. Broadband access and progress in ubiquitous computing have encouraged the progress of power line communication providing homogeneous

services via an authentic and strong system of the open system interconnection (OSI) model.

4.3.4 Implementation of Communication Models for IoT

While the idea of incorporating computers, sensors, and networks to observe and supervise devices has been around for decades, the recent convergence of crucial inclinations is unfolding in the current reality for the IoT. The main objective of IoT therefore is to provide a fully interconnected “smart” world, with cordial and good correlations between objects. With this, the communication models for IoT are given below.

Device-to-Device Communication Model: In device-to-device (D2D) communication model, communication is said to take place between two devices directly, instead of happening through an application server. D2D communication is used for local services, where users’ data is transmitted directly between the users without the aid of any network based on proximity service. Next, D2D model is used where emergency communication is concerned, in case of different natural disasters like hurricanes and earthquakes. This is because of the reason that the conventional communication network does not work due to the damage caused by the network. Hence, an ad-hoc network is said to be established through D2D. Finally, when IoT enhancement integrates with IoT, a precisely interconnected wireless network is said to be created. An example of D2D-based IoT is the communication between one vehicle and another in the Internet of Vehicles (IoV).

Device-to-Cloud Communication Model: Here, communication is said to take place between a device and cloud. Here, the IoT device is said to be connected via Internet cloud service for exchanging or transferring data and control messages.

Device-to-Gateway Communication Model: In device-to-gateway communication model, communication is said to take place between a device and the gateway node. Here, an application software acts as the intermediate between the device and the cloud server.

Back-End Data Sharing Communication Model: In this model, data is shared between objects via a back-end server.

< Prev   CONTENTS   Source   Next >