Deep Learning for Internet of Things Infrastructure

Uttam Ghosh, Mamoun Alazab, Ali Kashif Bashir, Al-Sakib Khan Pathan, Uttam Ghosh, Mamoun Alazab, Ali Kashif Bashir, Al-Sakib Khan Pathan

• Publisher: CRC Press
• Year: 2021
• Print ISBN:
  9780367457334
• eBook ISBN:
  9781000431957
• Topic: Computer Science
• Subtopic: Computer Networking
• Index: Computer Science

This book promotes and facilitates exchanges of research knowledge and findings across different disciplines on the design and investigation of deep learning (DL)–based data analytics of IoT (Internet of Things) infrastructures. Deep Learning for Internet of Things Infrastructure addresses emerging trends and issues on IoT systems and services across various application domains. The book investigates the challenges posed by the implementation of deep learning on IoT networking models and services. It provides fundamental theory, model, and methodology in interpreting, aggregating, processing, and analyzing data for intelligent DL-enabled IoT. The book also explores new functions and technologies to provide adaptive services and intelligent applications for different end users.

FEATURES

• Promotes and facilitates exchanges of research knowledge and findings across different disciplines on the design and investigation of DL-based data analytics of IoT infrastructures
• Addresses emerging trends and issues on IoT systems and services across various application domains
• Investigates the challenges posed by the implementation of deep learning on IoT networking models and services
• Provides fundamental theory, model, and methodology in interpreting, aggregating, processing, and analyzing data for intelligent DL-enabled IoT
• Explores new functions and technologies to provide adaptive services and intelligent applications for different end users

Uttam Ghosh is an Assistant Professor in the Department of Electrical Engineering and Computer Science, Vanderbilt University, Nashville, Tennessee, USA.

Mamoun Alazab is an Associate Professor in the College of Engineering, IT and Environment at Charles Darwin University, Australia.

Ali Kashif Bashir is a Senior Lecturer/Associate Professor and Program Leader of BSc (H) Computer Forensics and Security at the Department of Computing and Mathematics, Manchester Metropolitan University, United Kingdom.

Al-Sakib Khan Pathan is an Adjunct Professor of Computer Science and Engineering at the Independent University, Bangladesh.

1 Data Caching at Fog Nodes under IoT Networks

Review of Machine Learning Approaches

Riya and Nitin Gupta

National Institute of Technology, Hamirpur

Qin Xin

University of the Faroe Islands

Contents

1. 1.1 Introduction
   1. 1.1.1 Importance of Caching at Fog Nodes
2. 1.2 Applications of Data Caching at Fog Nodes for IoT Devices
3. 1.3 Life Cycle of Fog Data
4. 1.4 Machine Learning for Data Caching and Replacement
5. 1.5 Future Research Directions
6. 1.6 Conclusion
7. References

1.1 Introduction

In recent years, small devices embedded with sensors produce a large amount of data by sensing real-time information from the environment. The network of these devices communicating with each other is recognized as the Internet of Things (IoT), sometimes called as Internet of Everything [1]. The data produced by the IoT devices need to be delivered to the users using IoT applications after processing and analyzing. Further, data produced by the IoT devices are transient, which means that generated data have a particular lifetime, and after that lifetime, the data become useless and hence discarded [2]. Therefore, it is required to store the data somewhere near the IoT devices [3]. Simultaneously, suppose the data produced by the IoT devices are stored at the cloud server. In that case, it adds communication overhead, as the IoT users need to contact the cloud server whenever they require any data.

Fog computing is a decentralized approach to bring the advantages and intelligence of cloud computing such as storage, applications, and computing services near the end devices somewhere between the cloud and the end devices [4,5]. Fog nodes can be anything such as servers, networking devices (routers and gateways), cloudlets, and base stations. These nodes are aware of their geographical distribution as well as the logical location in the cluster. They can operate in a centralized or in a distributed manner and can also act as stand-alone devices. These nodes receive inputs from the data generators (IoT devices), process them, and provide temporary storage to the data. Fog nodes are intelligent devices that decide what data to store and what to send to the cloud for historical analysis. These devices can be either software or hardware, arranged in a hierarchy, and used to filter data transmitted by the sensors devices. These devices should have less latency, high response time, optimal bandwidth, optimal storage, and decision-making capability. At fog nodes, intelligent algorithms are embedded to store data, computing, and forward data between various layers. The member function of the fog node in the fog-cloud network is depicted in Figure 1.1. In this figure, the compute module is responsible for processing data and calculating the desired result. The storage module is responsible for storing data reliably so that robustness can be achieved. Further, various accelerator units such as digital signal processors and graphics processing units are used in critical tasks to provide additional power. In contrast, the network module is responsible for the guaranteed delivery of data.

FIGURE 1.1 Functions of fog nodes.

Fog computing only complements cloud computing by providing short-term analytics, unlike cloud computing which offers long-term analytics. However, it is to be mentioned that fog computing does not replace cloud computing [6]. There are prominent six characteristics that differentiate fog computing from other computing paradigms [7,8].

1. Awareness and Low Latency: Fog nodes are aware of their logical location in the whole system and offer very low latency and communication costs. Fog nodes are frequently placed near the edge devices, and hence they can return reply and other analysis much faster than the cloud nodes.
2. Heterogeneity: Fog nodes generally collect different forms of data and from other types of devices through various kinds of networks.
3. Adaptive: In many situations, fog computing deals with uncertain load patterns of various requests submitted by different IoT applications. Adaptive and scaling features of fog computing help it to deal with the scenario mentioned earlier.
4. Real-Time Interaction: Unlike cloud computing, which supports batch processing, fog computing supports real-time interaction. The real-time data, which is time -sensitive, is processed and stored at fog nodes and is sent back to the users whenever required. On the contrary, the data which is not time -sensitive and whose life cycle is long is sent to the cloud for processing.
5. Interoperability: Because fog computing supports real-time interaction, it requires the cooperation of various providers leading to the interoperable property of fog computing.
6. Geographically Distributed: Unlike a centralized cloud, the applications serviced by fog nodes are geographically distributed, like delivering seamless quality videos to the moving vehicles.

Further, the processing time of fog nodes is significantly less (millisecond to subsecond). This technique avoids the need for costly bandwidth and helps the cloud by handling the transient data. To facilitate fog computing, the node should exhibit autonomy (property to take decision independently without the intervention of other nodes), heterogeneity, manageability, and programmability. Figure 1.2 shows fog computing architecture where IoT devices are connected to fog nodes, and then fog nodes are further connected to the cloud nodes [9].

FIGURE 1.2 An architecture of fog computing.

The architecture of fog computing consists of three layers [10]:

1. Terminal Layer: This is the lowermost layer and consists of the IoT devices such as mobile phones and sensors, which detect the information from the environment by sensing it and then transmit the detected information to the upper layer. The information is transmitted in the form of data streams. The IoT data streams are the sequence of values emitted by the IoT devices or produced by one application module for another application module and sent to the higher layer for processing.
2. Fog Layer: This layer consists of various switches, portals, base stations, and specific servers. This layer lies between the IoT devices and the cloud and is used to process data near the IoT devices. If fog nodes cannot fulfill the terminal layer’s request, then the request is forwarded to the cloud layer.
3. Cloud Layer: This layer consists of high-performance servers used for the storage of data and performing powerful computing.

Generally, IoT devices do not have processing power and storage, due to which they suffer from many problems such as performance, reliability, and security [11]. Fog nodes are capable of performing operations that require a large number of resources on behalf of IoT devices which are generally resource-constrained. This makes end devices less complex and also reduces power consumption. Further, fog computing also supports real-time interactions between the IoT devices and fog nodes. The data is available to the IoT devices quickly, unlike cloud computing where batch processing is mostly used. IoT devices are resource-constrained and generally do not have security features for which fog nodes act like proxy servers and provide extra security features. Fog nodes regularly update the software and security credentials and check the safety status of these devices.

Fog computing also offers the implementation of various service models such as Software as a Service (SaaS), Platform as Service (PaaS), and Infrastructure as a Service (IaaS) [12,13]. Due to such advantages, various frameworks such as Google App Engine, Microsoft Azure, and Amazon Web Services using cloud computing have also started supporting fog computing for providing solutions to advanced distributed applications that are geographically dispersed and require low-latency computational resources. They are also using dedicated nodes with low-latency computational power, also called mist nodes (lightweight fog nodes), and are sometimes placed closer to the IoT devices than fog nodes [14,15]. Hence, the integration of IoT with fog computing brings many such advantages.

1.1.1 Importance of Caching at Fog Nodes

The IoT devices do not have to contact the remote server, i.e., cloud, whenever they require some data. The IoT devices first check data in the cache of fog nodes. If required data is present, then the fog nodes return the data to the IoT devices; otherwise, they contact the cloud for the needed data. Hence, caching of data at fog nodes reduces the transactional latency. Moreover, fog computing requires lesser bandwidth to transfer the data [16]. As fog computing supports hierarchical processing, the amount of the data needed to be transferred from the IoT devices to the clouds is more petite. In contrast, the amount of data transmitted per unit of time from the fog node to the IoT devices is more, which leads to improvement in overall throughput. Hence, caching data at fog nodes decreases the general operational expenses. Data is stored in the distributed manner at fog nodes which can be deployed anywhere according to the requirements.

Further, caching of data at fog nodes helps reduce load at the cloud servers as the data whose frequency of inte...

Table of contents

• Cover
• Half Title
• Title Page
• Copyright Page
• Dedication
• Table of Contents
• Acknowledgments
• Editors
• Contributors
• Chapter 1 Data Caching at Fog Nodes under IoT Networks: Review of Machine Learning Approaches
• Chapter 2 ECC-Based Privacy-Preserving Mechanisms Using Deep Learning for Industrial IoT: A State-of-the-Art Approaches
• Chapter 3 Contemporary Developments and Technologies in Deep Learning–Based IoT
• Chapter 4 Deep Learning–Assisted Vehicle Counting for Intersection and Traffic Management in Smart Cities
• Chapter 5 Toward Rapid Development and Deployment of Machine Learning Pipelines across Cloud-Edge
• Chapter 6 Category Identification Technique by a Semantic Feature Generation Algorithm
• Chapter 7 Role of Deep Learning Algorithms in Securing Internet of Things Applications
• Chapter 8 Deep Learning and IoT in Ophthalmology
• Chapter 9 Deep Learning in IoT-Based Healthcare Applications
• Chapter 10 Authentication and Access Control for IoT Devices and Its Applications
• Chapter 11 Deep Neural Network–Based Security Model for IoT Device Network
• Index