Cloud computing, Machine-to-Machine (M2M), Artificial Intelligent (AI), and Blockchain are among the large group of technologies that enables effective and reliable IIoT [52], with IoT forming the backbone. Figure 6 shows the structure of a typical IoT technology. The things here are regular objects, such as a VR headset, watch, electric cooker, parking lot or garage, security cameras, shopping cart, etc., that are made smart because of the sensor nodes embedded in them. This sensing ability enables these things to interact through a common infrastructure, the Internet, with a management center for decision making. The interaction is possible through the various wireless communication technologies available. For the protection of data, a firewall is designed to secure the information exchange between the objects and the center. Communication device manufacturers have provided solutions to protect this data. The network layer provides the platform on which different communication technologies can function. These include long-range licensed and unlicensed wireless bands. The licensed wireless technology, such as 2G/3G and LTE involves, the use of existing telecommunication infrastructures, while the unlicensed wireless (LoRa, ZigBee, etc.) is an ad hoc setup with Internet capability (through an Internet modem or a gateway). The smart objects, set up in this fashion, have various applications such as in homes, city environment, traffic, etc. Their management is also done through any internet-empowered device such as smartphones, laptops, tablets, etc. Figure 7 represents the application of IoT in the industrial environment, the IIoT. In industrial applications, the sensor nodes are embedded in the industrial equipment for monitoring, operations, and management. The communication technologies remain the same; however, the processes that are managed are of critical nature. Applications include oil and gas industries, chemical, automobile, food processing industries, etc. While IoT could be less stringent in terms of criticality, IIoT is critical with timing, reliability, privacy, and security of data being gathered in operation or monitoring.

IoT and IIoT are thought to have three or five architectural functionality layers ^[53][54][53,54]. Meanwhile, in [55], for a Man Like Nervous (MLN) system, other layers are involved. Three-layered IoT/IIoT architecture include perception, network, and application layers. While in five-layered architecture, in addition to the three layers are processing and business layers. However, in any type of architecture proposed or adopted, the network or transport [53] layer is responsible for the communication of devices among themselves and with the network infrastructure (e.g., gateway, server, etc.). IIoT allows the industries to meet their operational targets, among other things. However, most of the peculiar needs of the industries are centered on the capabilities of communication technologies as supported by the network layer. These designs need to include latency—time required for data propagation and processing, topology–device connection style, throughput—the volume of data that the network can support, scalability—number of supported devices in the network, the security of data and energy—IoT device operational time without power supply [56]. Table 3 lists the common types of communication technologies used by IoT/IIoT and the network-supported features.

To understand the WSN technology, a study of a typical sensor node is important. Figure 3 shows the four major units of a node: power, sensing, processing, and transceiver units. The power unit provides the energy required by other units in the node. This power source in most cases is provided by a battery. The sensing unit monitors the physical quantities that are presented in analog form or signals. With the aid of an Analog-to-Digital converter (ADC), the analog signal is converted to a digital signal. Both the sensor and ADC form the sensing unit. The digital signal is then processed by a micro-controller or processor. The processed data are stored temporarily in the storage section of the processing unit. The transceiver unit enables the node to connect with other nodes or the gateway in the network [42].

The network architecture of WSN is considered to be of two types: structured and unstructured [43]. A structured WSN network design provides for a well-mannered approach to its planning and distribution of the nodes. With this type of deployment, the network maintenance rate is minimal, which leads to a low management expenditure as well. While in an unstructured network, the nodes’ distribution and placements are scattered without proper planning. This type of design features dense and random deployment of nodes. This will lead to uncovered areas of deployment in the field, which may in-turn need monitoring. Since the nodes are not carefully placed, node failure detection and connectivity management become difficult to handle. Sensor nodes are of different types, which are designed or described based on the type of physical quantity they monitor. As shown in Figure 4, sensors such as humidity, temperature, flow, vibration, power sensors, etc., are connected wirelessly to a base station that is capable of aggregating data gathered from the sensors to the network server. The user computer connects with the server to access the data and further process the data, which will be presented in a format that is more understandable by the decision making team. The wireless media used by these nodes enables a multi-hop communication. These communication links could be provided by radio or infrared technologies. Where a radio link is used, WSN often uses the unlicensed frequency band known as the Industrial, Scientific, and Medical (ISM) band [42]. These frequency bands are specified by the International Table of Frequency Allocations, and they also provide huge spectrum allocation and are unregulated but highly prone to interference.

The description of WSN topology is associated with the expected capabilities of sensor nodes, which enable the deployment of several numbers of nodes in a sensor field [44]. While four major types of topologies can be adapted for field deployment, such as, star, tree, point-to-point, and mesh topologies [36], well-planned handling of these arrangements are necessary for maintenance purpose [42]. Without a proper topology arrangement, maintenance-related issues become difficult to handle. These issues usually arise in deployment phases: pre-deployment and deployment phase, post-deployment phase, and redeployment of more nodes phase [42]. In some cases, the types of WSN are classified based on the field of application, such as health, industry, military, agriculture, etc. [45]. However, WSN types can be identified as terrestrial, underground, underwater, mobile, and multimedia wireless sensor networks [46].

Although WSN technology has had performance improvements over the years [47], with many research questions getting resolved, other challenges are still present. Some of these challenges arise from node structure, while others are posed by the environment in which they operate in. Reference [48] highlighted some of these design challenges encountered by WSN. These include the operational environments, where WSN are deployed in areas that affect its operation, such as the industries, underwater, underground, etc.; the power consumed by nodes to process its data, transmit it, or receive any data; the scalability of these nodes, since they are deployed in large numbers for a specific application; provision of good quality of service; fault tolerance in the network; limitation in computational and storage capabilities; security of data in terms of confidentiality, integrity, and availability.

The network or transport layer handles the coverage range of IoT/IIoT. Each type of communication technology has an approximate range it can cover. Therefore, depending on the requirements needed for an application, appropriate technology can be employed. However, for IIoT where timing is critical, latency is an important feature that needs to be put into consideration to achieve reasonable reliability of deployment. To this end, a true picture or understanding of the industrial wireless communication for the application of IoT is necessary.

Some of the challenges faced by Industrial IoT are the requirements needed for their deployment in various settings. Although different fields of application have their requirements [21], most of these requirements are expected or provided by the network layer of the IoT architecture. For example, in [66], the scalability, throughput, latency, energy consumption, topology, and security of data were explained. More so, the interoperability of devices to support the heterogeneous nature of IoT is presented as a future requirement for IoT in [67]. Standardization, architecture, and the need for an increase in storage are some other challenges that IIoT faces [68]. This research paper presents the challenges that are related to the network layer.

Scalability: The capacity provided by the global information infrastructure to connect a large number of common objects describes the scalability. This challenge can present itself on different levels [69]. The addressing and identification of devices pose a challenge; however, this issue is mitigated by the use of IPv6 and 6LoWPAN protocols [70]. Due to the large size of these connected objects, a high volume of data communication (big data) and networking capabilities are issues to be tackled. The application layer also contributes to the challenge posed by scalability as a result of the numerous services and service execution options that are generated by heterogeneous connected IoT devices. Therefore, one of the challenges in scalability is the ease of adding new devices to the network without major interruption.

Latency: Since the industrial application of IoT is time-critical and usually deployed in environments that are noisy and hard on the wireless communications, the Quality of Service (QoS) provided by IIoT is often characterized by the real-time achievement of the deadline set by an end-to-end communication task, from sensing to control execution in the system [71]. In industries, various communication traffic is generated, and depending on the type, the latency and reliability provided by IIoT can either be least critical or most critical. These types of traffic are explained in [72]. Table 4 presents a summary of the latency and reliability levels needed for the communication traffic.

Industrial traffic (data) that requires fractions of seconds in latency in IIoT communications includes the emergency, critical control, and remote control traffics. This traffic is used for safety purposes, such as the leakage of radiation or poisonous gases in industries [72], continual flow of industrial processes such as automation processes, which require 1 ms latency rate [73], and control of unmanned vehicles, which require 50 ms latency and 1–10−8