manufacturing [153]. In addition, private cellular connections could help in assigning spectrum bands and base stations such that data transmission is performed via secure and reserved channels [154, 155]. A main differentiator between 5G and previous generations of cellular networks lies in its significant emphasis on MTC and IoT. For instance, 5G air interface has enacted novel techniques in PHY and MAC layers that accommodate MTC [156]. It seems that 5G leads to convergence of the many different communication technologies as it becomes the standard wireless technology.
5G specifically focuses on supporting communication at very low latencies, unparalleled reliability, and massive IoT connectivity. Based on these distinct features in 5G networks, the emerging diversified telecommunication services are categorized into three main classes [157]: (1) ultra-reliable and low latency communications (URLLC); (2) enhanced mobile broadband (eMBB); and (3) massive machine-type communications (mMTC).
1.5.4.1 Ultra-Reliable and Low Latency Communications
Compared with 4G network, E2E radio network latency in 5G is reduced to 1 ms with peak data transfer rate of 20 Gbps; this offers an ultra-responsive connection with ultra-low latency [158].15 URLLC is expected to play a key role in a wide range of mission-critical and industrial automation applications such as remote medical assistant, autonomous vehicle control, and robot and drone control that rely on high data rates and low latency [159, 160]. One important aspect of URLLC scenarios is to communicate in real time or within a very short time. Consequently, applications that require highly reliable communication can be implemented in 5G, such as remote-controlled plants or smart factories. At the same time, 5G exploits lower frequency bands that propagate farther in the environment, providing a more robust means of communication between IoT devices in buildings. Such a feature also leads to prolonged devices battery life (about years).
In addition to highly reliable communications [161], 5G new radio introduces the concept of extra transmission redundancy to better support industrial applications in URLLC scenarios [162]. Design of URLLC services heavily relies on a number of factors including more reliable channel coding techniques, effective resource sharing (control-flow and data), and grant-free transmission for uplink data, all supported by 5G new radio standard [163].
1.5.4.2 Enhanced Mobile Broadband
5G is expected to support around 29 billion devices connected to IoT by 2022 [164, 165]. eMBB delivers high data rates across 5G coverage area, where downlink data rates of at least 100 Mbps per device is supported over a typical dense urban environment. With the increased bandwidth, the high-performance connectivity is also sustained for indoor spaces such as industrial campuses [166]. Therefore, bandwidth hungry applications such as virtual reality for remote maintenance worker assistance and production line imaging for quality inspection significantly benefit from the higher bandwidth available through 5G eMBB. In addition to fast data rate, 5G eMBB provides low latency of 4 ms over the air, enabling real-time data transmission, processing, and decision-making for autonomous warehousing robots and industrial machine control [167]. One approach for providing eMBB is to implement millimeter wave (mm-wave) technology and install high-frequency mm-wave antennas [168].
1.5.4.3 Massive Machine Type Communication
The 5G solution that serves a large number of MTC applications is mMTC, which deals with scalable connectivity for massive number of devices. 5G offers wireless communication to over 1 million IoT terminals (with diverse QoS requirements) per square kilometer of area [169]. mMTC data communication is typically infrequent, making it ideal for alerting systems or periodic sampling (e.g., indicating manufacturing equipment failure and low frequency environmental sensing). This feature is suitable for a wide range of use cases across utilities, industrial campuses, and logistics. mMTC should handle different exceptional challenges, such as varied and intermittent traffic, QoS provisioning, large signaling overhead, and congestion in RAN [137].
1.6 Wireless System Design Enablers and Metrics for Emerging IIoT Applications
IIoT is one of the key components of the Fourth Industrial Revolution. It provides customized architectures and standardized interfaces for data acquisition, transmission, and analytics in industrial applications [25]. Diverse industrial applications differ in terms of operational settings, technical requirements, and service environments. Therefore, it is not possible to provide one multi-purpose wireless solution for all IIoT use cases. Each wireless system design requires theoretical and experimental measures based on its expected performance. In this section, we first review conventional technical enablers in design of wireless networks for IIoT and then discuss the metrics on the desired performance.
1.6.1 General Technical Enablers in Design of Wireless Network for IIoT
The National Institute of Standards and Technology (NIST) proposed a reference framework as a guideline that helps users select and design a given wireless system, customize its configuration based on the specific application requirements, successfully deploy it, and finally ensure its performance for industrial environments [55]. Based on this framework, the design of industrial wireless communication should be evaluated from three major aspects: system modeling and verification, radio resource management (RRM) schemes, and protocol interfaces design.
1.6.1.1 System Modeling and Verification
The connectivity in IIoT systems could exploit well-established communication protocols to reduce network configurations and customizations. However, the increasing integration of communication into automation aspects makes IIoT systems more complex and prone to errors (e.g., device failures, mistakes in configuration). Given that failure in communication may be catastrophic in industrial applications using IIoT, it is essential to use proper system models and verification schemes to increase the level of certainty in IIoT systems. There are different approaches to create system models, such as theoretical inference and simulation tests.
An important reference in IIoT wireless system design is modeling data traffic patterns to capture and predict the dynamic behavior of systems and handle system complexity. However, we initially need to understand the theoretical basis of inference models and the conditions underlying their effectiveness before choosing the method apt to IIoT environments and service characteristics. This justifies a rising need for network simulation platforms and testbeds that emulate real-world industrial systems and perform system-level verification [170]. The simulation frameworks could also assist in performance evaluation of wireless networks for the next-generation factories and process automation systems [171].
1.6.1.2 Radio Resource Management
The rise of ultra-dense and dynamic wireless networks in the Industry 4.0 paradigm implies a further number of simultaneous transmissions. Therefore, it is necessary to efficiently utilize limited wireless resources such as RF spectrum resources and radio network infrastructure to fulfill strict QoS requirements and achieve a reasonable level of performance across IIoT systems. Radio resource management (RRM) involves strategies and algorithms that manage radio transmission for reliable service delivery in dynamic and diverse wireless networks.
The fundamental challenge in IIoT wireless communication is cross-technology interference combined with harsh signal propagation conditions in industrial systems. This results in deficient networks performance and service failures regardless of prudent initiatives [172]. One possible solution for interference mitigation is coexistence mechanisms. There are two principal concerns in the design of coexistence mechanisms: interference management and load balancing [173]. Since RRM includes transmission power management, radio resources scheduling, user allocation, and preventive–reactive congestion control, effective RRM procedures could be exploited in coexistence mechanisms to mitigate the interference level. Another possible approach
