Industrial Internet of Things and Its Applications: Concepts, Architecture, Key Technologies, Applications, and Challenges

The Internet of Things (IoT) is changing the way people interact with things around them. It is not only attractive to consumer applications, but also to the industrial sector, which is Industrial Internet of Things (IIoT).

IIoT connects all industrial assets, including machines, control systems, information systems, and business processes. Therefore, a large amount of data is collected to provide analytical solutions for achieving optimal industrial operations. IIoT affects the entire industry value chain and is a necessary requirement for intelligent manufacturing.

The Internet of Things is an extension of the Internet, and conceptually, it cannot be separated from the Internet. The Internet of Things is based on the Internet, with the former centered around perception and the latter centered around links. In fact, the traditional internet is evolving from a simple "human internet" to an internet that includes both people and things. There is reason to believe that the boundary between the Internet of Things and the Internet will become increasingly blurred.

The Industrial Internet can be seen as a subset of the Internet. It is the "industrial interconnection" within enterprises, between enterprises, product design, production, marketing, and services. And the Industrial Internet of Things can be seen as a subset of the Industrial Internet, with perception (or sensing) as its basic feature. In fact, in layman's terms (although inaccurate but easy to understand), the Industrial Internet focuses on "industrial management", while the Industrial Internet of Things is the "industrial operation" in "industrial management".

However, in some cases, people do not pay much attention to distinguishing between the Industrial Internet of Things and the Industrial Internet of Things.

1. The Internet of Things and Industrial Internet of Things 1.1 The Internet of Things (IoT) can achieve the connection of things, people, and people at any time and place. The Internet of Things enables sensors and devices to seamlessly communicate in intelligent environments and facilitate cross platform information sharing in a convenient manner.

In the past few years, the Internet of Things has become a new trend, and data collection devices in crop networking can be used for mobile devices, transportation facilities, public facilities, and household appliances. Electronic devices (such as watches) and household appliances (such as refrigerators) in daily life can be connected to the Internet of Things network for remote control. In the Internet of Things, sensor devices perceive and transmit data, and devices and objects can be connected through various communication methods, such as Bluetooth, WiFi, ZigBee, and GSM. These devices receive commands from remote control devices, thus achieving the integration of computer systems and the physical world, and improving people's living standards.

1.2 Industrial Internet of Things (IIoT) covers the field of industrial communication technology for machine to machine (M2M) and automation applications. IIoT enables people to better understand industrial production processes, thereby achieving efficient and sustainable production. The Industrial Internet aims to achieve intelligent collaboration of massive industrial entities and change the future industrial infrastructure of industrial production forms. It requires the use of new generation technological concepts to model and control different types of industrial entities and even the entire industrial network, efficiently integrate industrial and social resources, and achieve intelligent development of industrial entities.

1.3 Differences and connections between the Internet of Things and the Industrial Internet of Things. Overall, IIoT can be regarded as a subset of IoT.

(1) Different types of services. Typically, the Internet of Things is still human centered, and "things" refer to the interaction and interconnection between intelligent electronic devices to increase human perception and response to the surrounding environment. Generally speaking, IoT communication can be divided into two categories: machine to user communication and client server interaction.

The communication in IIoT is machine oriented and can span various markets and applications. The IIoT scenarios include: i) monitoring applications, such as factory production process monitoring, and ii) creative applications for self-organizing systems, such as automated industrial factories.

(2) The Internet of Things places more emphasis on designing new standards to connect new devices to the Internet ecosystem in a flexible and user-friendly way, with different devices being connected. In contrast, the current IIoT design emphasizes the integration and connection of factories and machines to provide more efficient production and new services. For this reason, compared to the Internet of Things, IIoT can be more of an evolution than a revolution.

(3) The network requires different IoT to be more flexible, allowing for temporary and mobile network structures with lower timing and reliability requirements (except for medical applications). On the other hand, IIoT typically uses a fixed network structure with fixed nodes and centralized network management. The communication of IIoT is a machine to machine connection that must meet strict real-time and reliability requirements.

(4) The amount of data generated by the Internet of Things varies, and therefore, the amount of data transmitted is moderate or large. Currently, IIoT is mainly focused on big data analysis, such as predicting industrial maintenance. Therefore, the amount of data transmitted in IIoT is very large.

2. We elaborate on the architecture of industrial Internet of Things from different perspectives in order to gain a deeper understanding of its essence.

2.1 IIoT architecture description based on system concepts. From the perspective of industrial systems, reference [4] provides a relatively concise IIoT system structure, which is divided into three layers: physical layer, communication layer, and application layer.

Figure 1 Industrial Internet of Things System Structure [4]

physical layer

Composed of widely deployed physical devices such as sensors, actuators, manufacturing equipment, facility utilities, and other industrial manufacturing and automation related objects.

Communication layer

The integration of numerous communication networks, such as wireless sensor and actuator networks (WSANs), 5G, M2M, SDN, etc. In intelligent industrial applications, various network technologies will inevitably support the interconnection of a considerable number of sensors and actuators.

application layer

Composed of various industrial applications, including smart factories, smart supply chains, etc. These intelligent industrial applications utilize numerous sensors and actuators to achieve real-time monitoring, precise control, and effective management.

The above structure is simple and clear, making it easy to understand the basic concepts of IIoT.

2.2 Description of Service Oriented IoT Architecture Reference [2] describes the Service Oriented (SOA) IoT architecture, which emphasizes scalability, modularity, and interoperability between heterogeneous devices using different technologies, facilitating a technical understanding of IIoT. This structure is divided into four layers: sensor layer, network layer, service layer, and interface layer.

Figure 2 Service oriented IoT architecture [2]

Sensing layer

The sensing layer is a fundamental feature of the Internet of Things, consisting of sensors (such as RFID and smart sensors) and corresponding data perception/collection protocols.

The Internet of Things can be seen as a network that enables remote connection and control on a global scale. The basic requirement for remote connection and control of objects is sensing. At the sensing layer, wireless intelligent systems with tags or sensors can automatically perceive and exchange information between different devices. These technologies have significantly improved the ability of the Internet of Things to perceive and recognize objects or environments. In some industries, it is necessary to assign a Universal Unique identifier (UUID) to every possible service or device, and devices with UUID can be easily identified and retrieved.

network layer

The role of the network layer is to connect all objects and allow devices to share information with other connected devices. The network layer is actually a heterogeneous network composed of multiple networks.

The network layer can obtain information from existing IT infrastructure (such as commercial systems, transportation systems, power grids, healthcare systems, ICT systems, etc.) and aggregate information. In IIoT, the devices providing services are usually deployed in heterogeneous networks, and all related devices are introduced into the service network. This process may involve QoS management and meeting the needs of user applications. On the other hand, for dynamically changing networks, automatic discovery of networks is crucial. The device needs to automatically assign roles to deploy, manage and schedule, and switch to any other roles at any time as required. These features enable devices to collaborate to complete tasks. In order to design the network layer of IIoT, designers need to address issues such as network management technology for heterogeneous networks, energy efficiency, QoS requirements, service discovery and retrieval, data and signal processing, security and privacy, and so on.

Service layer

The service layer relies on middleware technology, which provides seamless integration of IoT services and applications. Middleware technology provides a low-cost platform for the Internet of Things, where hardware and software platforms can be reused. The service layer mainly provides a service specification for middleware developed by different organizations. A good service layer will be able to identify common application requirements and provide APIs and protocols to support and meet the required services, applications, and user needs. The service layer also handles all service-oriented issues, including information exchange and storage, data management, search engines, and communication.

Interface layer

In IIoT, most devices are manufactured by different manufacturers/suppliers, so they may follow different standards/protocols. Due to heterogeneity, there are many interaction issues in information exchange, communication between devices, and collaborative event processing between different devices. In addition, the increasing number of devices participating in the Internet of Things makes dynamic connection, communication, disconnection, and operation more difficult. The interface layer also needs to simplify the management and interconnection of things. Universal Plug and Play (UPnP) defines a specification for facilitating the interaction of various services provided. Interface profiles are used to describe the specifications between applications and services.

2.3 Description of Cloud based IIoT Infrastructure Reference [5] describes a typical cloud based IIoT infrastructure consisting of three layers: device layer, gateway layer, and cloud service layer, as shown in Figure 3.

Figure 3 Cloud based Industrial Internet of Things Architecture [5]

Device layer

The device layer includes heterogeneous IIoT devices, ranging from powerful computing units to extremely low-power microcontrollers. These devices are connected to the gateway layer through various wired and wireless networks.

Most IoT devices have limited resources, including memory size, computing power, and communication bandwidth. In addition, these devices and the network technologies they use have a high degree of heterogeneity, which poses great challenges to the interconnection of IIoT devices. Due to heterogeneity, interoperability between devices should be prioritized so that heterogeneous devices can be transformed in a syntax and semantics acceptable to users.

Gateway layer

Most companies and organizations deploy their own custom gateways to manage local IIoT networks, aggregate data, and act as bridges to the cloud. These customized gateways are typically part of the deployed IIoT infrastructure, leading directly to a "chimney like" solution. This further leads to interoperability issues, where data and services provided by one organization cannot be shared or used by devices of other organizations (due to different network protocols, data formats, etc.), and the security mechanisms used are usually proprietary.

Cloud service layer

The cloud service layer provides cloud related functions, such as database services and application services, to manage the data provided by this Internet of Things. The local Internet of Things and cloud service layer together constitute the most common existing cloud based Internet of Things infrastructure.

3. Key Technologies of Industrial Internet of Things

3.1 Identification and Tracking Technology Industrial Internet of Things involves identification and tracking technologies such as RFID systems, barcodes, and intelligent sensors. A simple RFID system consists of an RFID reader and an RFID tag. RFID systems have the ability to identify, track, and track devices, and are increasingly being applied in industries such as logistics, supply chain management, and medical service monitoring. The RFID system can also provide real-time information about the equipment involved, reduce labor costs, simplify business processes, improve the accuracy of inventory information, and improve business efficiency. There is still great room for development in RFID based applications.

In order to further promote the development of RFID technology, RFID can be combined with wireless sensor networks to better track objects in real-time. Especially the emerging wireless intelligent sensor technologies, such as electromagnetic sensors, biosensors, onboard sensors, sensor tags, independent tags, and sensor devices, further promote the implementation of industrial services. Integrating the data obtained from smart sensors with RFID data can create IoT applications that are more suitable for industrial environments.

3.2 Communication Technology IIoT involves various heterogeneous networks such as wireless sensor networks, wireless mesh networks, and wireless local area networks. These networks help industrial IoT exchange information. Gateways can facilitate communication between various devices on the network, and can also be used to handle complex nodes involved in communication on the network. Different devices may have different QoS requirements, such as performance, energy efficiency, and security. Many devices require batteries, so reducing the energy consumption of these devices is a top priority. The industrial Internet of Things also involves the use of existing Internet protocols, mainly communication protocols and standards such as RFID, NFC, IEEE802.11 (WLAN), IEEE 802.15.4 (ZigBee), IEEE 802.15.1 (Bluetooth), multi hop wireless sensor networks, machine to machine (M2M), traditional IP technologies such as IP, IPv6, etc.

3.3 Wireless networks have many cross layer protocols, such as wireless sensors or Ad Hoc networks (AHNs). Due to the different communication and computing capabilities of devices in the industrial internet, as well as different QoS requirements, some devices need to be modified before being applied to the industrial internet. In contrast, nodes in wireless sensor networks typically have the same hardware and network communication requirements, so there is no need to change them. In addition, the industrial Internet of Things utilizes the Internet to support information exchange and achieve data communication, while wireless sensor networks and AHNs do not require the use of the Internet for communication.

3.4 Service Management Industrial Internet Service Management refers to the management of high-quality IoT services to meet the needs of users or applications. The OSGi platform is a great example of using a dynamic SOA (Service Oriented) architecture to support the deployment of intelligent services. OSGi (Open Services Gateway Initiative), as an effective modular platform for service deployment, is widely used in various environments such as mobile applications, plugins, and application servers. In the industrial Internet of Things, service composition based on the OSGi platform can be implemented by Apache Felix iPoJo. Services can be divided into two types: primary services and secondary services. In a service-oriented industrial Internet of Things, services can be created and deployed by following the following steps: 1) developing a service composition platform; 2) The functions and communication capabilities of abstract devices; 3) Provide a common set of services. Service identity management includes context management and object classification. The industrial internet can also build a mirror for every real object in the network. The industrial internet also has a service-oriented and context aware architecture, where each virtual and physical object can communicate with each other.

4. 4.1 Application Overview of Industrial Internet of Things 4.1 IIoT Application Overview Industrial Internet of Things is an important branch of the Internet of Things, which has a wide range of applications, especially in the fields of energy, transportation (railways and stations, airports, ports), manufacturing (mining, oil and gas, supply chain, production), etc. It will play an important role.

When it comes to the application of industrial Internet of Things, there are two points worth noting:

One is that industrial applications similar to the concept of the Internet of Things have a long history. For example, process control and automation systems, industrial Ethernet and wireless local area networks (WALNs), programmable logic controllers (PLCs), wireless sensors, and radio frequency identification technology tags (RFID). But these applications are mainly based on "automation" considerations and are not externally connected to the enterprise.

The second is that the concept and technology of industrial Internet of Things can be extended and applied to some "non industrial industries", such as health, security, transportation, and so on.

In the "White Paper on Industrial Internet of Things" (2017 edition, China Institute of Electronic Technology Standards) [5], four application projects of industrial Internet of Things are described, including: leasing application based on machine tool Internet of Things, new insurance mode based on industrial Internet of Things, full automatic control of roll flow between grinding rolls, and logistics automation based on industrial Internet of Things. But the white paper does not provide a more comprehensive application description of IIoT.

The application classification of IIoT in industrial IoT can be divided into process automation (PA) applications and factory automation (FA) applications [4].

4.2.1 Process Automation

Process automation is characterized by an industrial process, such as the "autonomous" processes in chemical, petroleum, and power plants, which achieve control and management with little or no human intervention. Process automation systems typically integrate sensors, controllers, and actuators to achieve information collection, interaction, and process driving.

Power System Automation (PSA) is an example of process automation, which aims to automatically control, monitor, and protect power generation, distribution, and user systems through various instruments and control devices. PSA consists of three key components: data collection, remote monitoring, and control. Data collection is used to collect data from various sensing and control devices, which can be processed locally or in data centers; Remote monitoring is used to monitor the status of the power system, alert the operation center of abnormal situations, and prevent power outages; PSA control plays a role in operating the power system, controlling the substation from the operation center.

4.2.2 Factory Automation

Factory automation, also known as manufacturing process automation, utilizes robot systems and assembly line machinery to improve production capacity and efficiency. For example, the appearance and driving force of a robotic arm are usually similar to those of humans, and it can perform functions as a human operator, but it has stronger robustness, productivity, accuracy, and efficiency. Under the same goal, assembly lines typically break down complex tasks into smaller sub tasks and perform step-by-step operations according to the designed workflow.

4.2.3 Basic performance characteristics of IIoT

The performance requirements of industrial IoT are significantly different from those of consumer oriented IoT, especially in terms of time, scale, and reliability. These three basic performance characteristics can be described in terms of cycle time, number of nodes, and reliability.

Cycle time: The time required to receive commands from the control center and send sensor data to the control center. The cycle time depends on different applications. For example, the cycle time for general process automation is about a few hundred milliseconds.

Number of nodes: The number of nodes covered by a controller in the workspace, which represents the size of the system.

Reliability: Its characteristic lies in the quality of information transmission, which can be measured using packet error rate (PER). Obviously, IIoT requires high reliability, for example, for some industrial environments, PER10-9 is required.

In addition to the obvious application of factory automation, this section briefly introduces the application of industrial Internet of Things in several industries.

4.3.1 Smart Grid

The power grid includes three basic functions: the generation (generation), transmission, and distribution of electrical energy. For traditional power grids, due to many factors such as inefficient consumer appliances and lack of intelligent technology, inefficient energy transmission routing and distribution, unreliable communication and monitoring, especially the lack of energy storage mechanisms, there is huge energy waste in the power grid. In addition, the power grid also faces other challenges, including increasing energy demand, reliability, safety, emerging renewable energy, and aging infrastructure issues. To address the aforementioned issues, a smart grid based on information processing and communication technology has emerged, and the industrial Internet of Things will play an important role in the smart grid. IIoT will greatly improve the efficiency and safety of smart grids by providing intelligent devices or IoT devices (such as sensors, actuators, and smart instruments) for monitoring, tracking, analysis, and control.

4.3.2 Transportation

The industrial Internet of Things will play an increasingly important role in the fields of transportation and logistics. With more and more objects equipped with barcodes, RFID tags, or sensors, transportation and logistics companies can achieve real-time monitoring of mobile transportation from the origin to the entire supply chain. In addition, IIoT is expected to provide promising solutions for the transformation of transportation systems and the automotive manufacturing industry. Industrial Internet technology can track the current location of traffic vehicles, monitor their movement, and predict their future location. Other applications, such as using unmanned maritime vehicles to monitor seabed conditions, crossing the ocean to collect data, and so on.

4.3.3 Mining safety production

Due to the working conditions of underground mines, mine safety issues are receiving increasing attention. In order to prevent and reduce the occurrence of mining accidents, it is necessary to use IIoT to perceive mining disaster signals, so as to carry out disaster prediction and early warning, and improve the safety level of underground production. By using RFID and wireless communication technology on the surface and underground, it is possible to track the location of underground miners and analyze key data collected from sensors to enhance security measures. In addition, chemical and biological sensors can be used for early disease detection and diagnosis of underground miners. These chemical and biological sensors can extract information and biological information from the human body and organs, and detect harmful dust, gases, and other environmental hazards that can cause accidents.

4.3.4 Food supply chain

Due to the large geographical and temporal scale and complex operational processes of the food supply chain FSC, the food supply chain is characterized by dispersion and complexity. This complexity has brought many problems to food quality management, operational efficiency, and public food safety. IIoT can accurately track the entire process of food production, processing, storage, distribution, and consumption. The future FSC system will be more secure, efficient, and sustainable. A typical FSC industrial IoT solution consists of three parts: on-site devices, such as WSN nodes, RFID readers/tags, user interface terminals, etc; Backbone systems such as databases, servers, and various software and small computers connected by distributed computer networks; Wireless LAN, cellular network, satellite, power line communication, Ethernet and other communication infrastructure. Due to the ubiquitous networking capabilities provided by industrial internet systems, all these elements can be distributed throughout the entire FSC. In addition, IIoT also provides effective sensing functions to track and monitor the grain production process.

4.3.5 Health services

Based on the ubiquitous recognition, sensing, and communication capabilities in the industrial Internet of Things, all objects in the medical system (people, devices, drugs, etc.) can be tracked and continuously monitored. Through global connectivity, all healthcare related information, including logistics, diagnosis, treatment, rehabilitation, medication, management, finance, and even daily activities, can be effectively collected, managed, and shared. For example, a patient's heart rate can be collected by a sensor and then sent to the doctor's office. By using personal terminals and mobile internet access, medical services based on the industrial Internet of Things can be more mobile and personalized.

5. The challenges faced by the Industrial Internet of Things 5.1 Energy Efficiency Many IIoT application devices rely on batteries for long-lasting operation, and it is difficult or even impossible to replace batteries in many application scenarios, or the cost of replacing batteries is too high. In order to avoid the need to replace batteries during their service life, energy-saving design requirements must be put forward. In addition, IIoT application devices typically require intensive deployment, and the sensed data needs to be sent in the form of queries or in a continuous manner, which consumes a significant amount of energy. In summary, reducing power consumption and operating costs in IIoT is an important issue. LPWAN is an effective way to solve low energy consumption. It adopts multiple energy-saving designs to achieve low-power operation: usually forming a star topology structure, effectively avoiding the energy consumption of multiple packet routing; Transfer complexity to the gateway; Using narrowband channels to reduce noise levels and expand transmission range.

In addition to some methods to improve energy efficiency, such as using lightweight communication protocols or using low-power transceivers as described above, a new technological trend is energy harvesting. In fact, energy can be obtained from environmental resources, such as thermal energy, solar energy, vibration, and radio frequency (RF) energy, among others.

5.2 Real time IIoT devices are typically deployed in noisy environments with strict time and reliability requirements to collect environmental data and make control decisions in a timely manner. In IIoT, time slot scheduling is crucial for the QoS required by the network. For example, many industrial IoT networks achieve real-time communication through static data link layer scheduling for network resource management. However, this process is slow, not scalable, and will incur significant network overhead. However, the explosive growth of IIoT applications, especially in terms of its scale and complexity, has greatly increased the difficulty level of real-time requirements.

Most existing processing methods are built on a centralized architecture, with limited scalability. Recently, some hybrid and fully distributed IIoT resource management methods have been proposed. However, how to ensure limited response time to handle concurrent interference remains an urgent issue that needs to be addressed.

5.3 Coexistence With the significant growth of IIoT devices, many devices use similar spectra, which brings about coexistence problems. Temporary frequency interference between devices must be resolved. In addition, due to the dense equipment and large scale, the characteristics of each technology may bring additional challenges to IIoT technology.

In order to coexist, it is best for future IIoT devices to be able to detect, classify, and reduce external interference. Although there are some techniques for spectrum sensing and interference suppression, the time sampling window is long and requires large memory. Some efficient anti-interference technologies in conventional communication systems, although having good performance, are generally more complex and have high implementation costs.

The issue of device diversity in IIoT can be addressed from three dimensions: multimode RF, software flexibility, and cross technology communication. Multimode RF allows different IIoT devices to communicate with each other; Software flexibility supports multiple protocols, connection frameworks, and cloud services. Therefore, in the future, it is necessary to study how to achieve cross technology communication in IIoT devices.

5.4 Security is also a key issue of IIoT. Generally speaking, IIoT is a resource limited communication network that heavily relies on narrowband communication. Therefore, traditional protection mechanisms are insufficient to protect complex IIoT systems, such as security protocols, lightweight cryptography, and privacy protection. To ensure the security of IIoT infrastructure, industrial wireless sensor network encryption technology can be applied before applying the IIoT security protocol. However, some encryption algorithms have high resource requirements, such as public key cryptography (PKC). This issue is more prominent in the application of real-time demand for massive data exchange.

When designing secure IIoT infrastructure, the following security features need to be considered:

1) IIoT devices need to have tamper proof capabilities to resist potential physical attacks;

2) The storage of IIoT devices requires data encryption to achieve confidentiality purposes;

3) Confidentiality and integrity protection of communication networks between IIoT devices;

4) IIoT infrastructure requires efficient identification and authorization mechanisms, and only authorized entities can access IIoT resources;

5) Even if the device is physically damaged by malicious users, the system can still operate normally to ensure the robustness of IIoT.

Symmetric key cryptography can provide a lightweight solution for IIoT devices. However, if symmetric key encryption is used, especially for low capacity devices, key storage and key management are major issues. In addition, if a device in IIoT is compromised, it may leak all other keys. Public key cryptography typically provides more secure features and lower storage requirements, but the high computational overhead caused by complex encryption algorithms is also a major problem. Elliptic Curve Cryptography (ECC) is a good solution that provides smaller keys and reduces storage and transmission requirements.

The data of IIoT devices should follow specific patterns and rules for exchanging/publishing authentication. Although public key cryptography provides authentication and authorization schemes, it does not provide a global root authentication authority, which greatly hinders many practical deployment solutions. If a global root CA is not provided, designing a secure authentication system in IIoT would be very challenging. Therefore, if one intends to provide security certification for IIoT devices, a costly solution must be used, which conflicts with the main goal of the lightweight principle of IIoT. In addition, issuing certificates for each object in IIoT will be a huge challenge due to the large total number of objects.

5.5 Privacy is a very broad and diverse concept. The privacy in IIoT mainly faces two challenges: the data collection process and the data anonymization process. Due to limitations in information collection and storage, privacy protection can be ensured during the data collection process. However, considering the diversity of data anonymization, different encryption schemes may be adopted, which poses a challenge to privacy protection. In addition, the collected information needs to be shared between IIoT devices, so the computation of encrypted data is another challenge for data anonymization.

Outlook: The Internet of Things is often considered a disruptive technology to solve many of today's problems, such as smart cities, intelligent transportation, pollution monitoring, connected healthcare, and so on. As a subset of the Internet of Things, Industrial Internet of Things IIpT has opened up new avenues for intelligent processes and intelligent manufacturing. IIoT covers information collection, transmission and processing, as well as automation. In the future, the integration of artificial intelligence, big data, and blockchain will further promote the development of IIoT, thereby achieving more efficient, secure, and sustainable production and management.

Reference [1] Xia Zhijie, Architecture and Key Technologies of Industrial Internet, China Mechanical Engineering, 2018,29 (10): 1248-1259

[2] Li Da Xu, WuHe, Shancang Li, Internet of Things in Industries: A Survey, IEEE Transactions Industrial Informatics, volume.10, No.4, November 2014

[3] Emiliano Sisini, Abusayeed Saifullah, et al., Industrial Internet of Things: Challenges, Opportunities, and Directions, IEEE Transactions on Industrial Informatics, volume.14, April 2018

[4] Hansong Xu, WeYu, et al., A Survey on Industrial Internet of Things: A Cyber Physical Systems Perspective, IEEE Access, Volume 6, 2018

[5] Gang Wang, Zhijie Jerry, et al., Shi_ ChainSplitter: Towards Blockchain based IndustrialIoT Architecture for Supporting Hierarchical Storage, 2019 IEEE International Conference on Blockchain (Blockchain)

[6] White Paper on Industrial Internet of Things (2017 Edition), China Electronics Technology Standards Research Institute, September 2017

[7] Yasir Saleem, Noel Crespi, et al., Internet of Things Aided Smart Grid: Technologies, Architectures, Applications, Prototypes, and Future Research Directions, IEEE Access, Volume 7, 2019