Industry 4.0 Vision for the Supply of Energy and Materials. Группа авторов

algorithms, a unique physical frequency is dedicated to every node to offer a collision-free protocol. The FDMA-based protocols support multiple frequencies and require more costly hardware. They generally are not useful for IoT systems because of a high level of power consumption and more complicated design [193].

       CDMA: A MAC channel access method that enables transmission of multiple signals in a single transmission channel. A combination of special encoding scheme and spreading spectrum technology is exploited to send multiple signals through a single channel. The basic principle is that users have access to the whole bandwidth for the entire duration, but they utilize different CDMA codes; this assists the receiver to distinguish among different users. Given that the entire bandwidth is allocated to a CDMA channel, this scheme suffers from limited flexibility in adapting bandwidth, particularly for M2M communication in IIoT systems.

       OFDMA: A multiple access scheme that divides the entire channel resources into small time-frequency resource units. Since the available bandwidth is divided into multiple mutually orthogonal narrowband sub-carriers, several users could share these sub-carriers and simultaneously transmit data. In other words, the signal is first split into multiple smaller sub-signals, and resource units are allocated to them. Then, each data stream is modulated and transmitted through the assigned resource units. OFDMA allows several users with various bandwidth requirements simultaneously to transmit data at different (orthogonal) frequencies. Therefore, channel resources can be assigned with much more flexibility for different types of traffic. In addition to high spectral efficiency, OFDMA can effectively overcome interference and frequency- selective fading caused by multipath. OFDMA is a promising multiple access scheme adopted for wide range of mobile broadband wireless networks such as LTE, Wi-Fi6, and 5G [194–196].

      1.7.2 Contention-Based Schemes

In contention-based protocols, nodes perform random access competition with each other to access medium on demand. Before transmission, nodes perform channel sensing to verify whether the medium is clear and wait for a specified backoff period to transmit data. Each node performs the channel sensing independently, and the allocated channel will be accessible for the required duration. Once data communication is complete, the occupied channel is released. Compared with scheduled-based protocols, this class of protocols does not require centralized control and precise time synchronization [197]. Moreover, these protocols are adequately simple, adaptive toward change in network topology, and robust to variation in nodes traffic load and density [198]. They also do not require extra message exchange overhead, thanks to the independent decision-making process for channel access. There are several protocols under this class, such as ALOHA [199] and variants of CSMA [200]. BREATH is a self-adapting CSMA MAC protocol that provides reliable, energy efficient, and timely data transmission for industrial control applications [52].

A consequential drawback of contention-based schemes is that the probability of collisions and idle listening grows with increased node density, leading to unacceptable performance in terms of latency. To alleviate the effect of collisions, additional control packets may be added to the MAC protocol, resulting in noticeable control overhead in IIoT systems. Contention-based MAC protocols could be performed through synchronous and asynchronous protocols.

       Synchronous protocols: This class of schemes employs local time synchronization between nodes to alternately switch their operation mode between active and sleep modes. In these protocols, a node operates in active mode for packets listening or sleeping mode to decrease overhearing and idle listening. To prevent overload from frequent synchronization messages, the protocol could use infrequent synchronization, although it may decrease network adaptability to nodes mobility [189].

       Asynchronous protocols: Unlike synchronous protocols, this method does not require explicit scheduling between nodes. Instead, a low power listening (LPL) concept could be employed, where each node transmits data with a long enough preamble so that receiver is guaranteed to wake up during preamble transmission [201]. Basically, the receiver is often in sleep mode and wakes up shortly to sense the channel for every preamble. If a sender has data, it will send preamble to the receiver until it is awake and properly acquires the preamble. Then, the receiver remains in active mode to receive incoming data. After the transmission or reception period, all nodes check their data queue before going to sleep mode. Duty cycle and idle listening of asynchronous protocols could be decreased through dynamic preamble sampling [202]. The advantages of asynchronous protocols are flexibility to topology changes, less synchronization overhead, and a reduction in a receiver’s idle listening. Nevertheless, asynchronous protocols suffer from transmitters’ overemission before sending data, extra power consumption in unintentional receivers, and increased latency [189]. It also does not fully resolve the issue of channel collisions.

      1.7.3 Hybrid Schemes

The combination of scheduled- and contention-based techniques is adopted in the so-called hybrid MAC schemes. In this class of schemes, time is divided into two periods, and, based on the requirements, it cooperatively switches between contention- or scheduled-based techniques to serve both modes. During contention period, nodes access to the medium and utilize contention approaches to broadcast in a common broadcast frequency. Then, during schedule-based period, each node performs reliable unicast transmission. Random–distributed scheduling algorithms are exploited in base stations to allocate the schedule of each node. For IoT use cases, some hybrid schemes jointly integrate FDMA with contention-based protocols and TDMA. Several MAC protocols are proposed in the literature based on such hybrid schemes [203, 204].

      1.8 Smart Sensors

Similar to industrial development, sensors and instrumentation development could be classified into four categories: mechanical indicators, electrical sensors, electronic sensors, and smart sensors [205]. Recent evolutions and advancements of sensing technology in the field of Industry 4.0 are exclusively entitled as Sensor 4.0 [205].

      Smart sensors are advanced platforms often associated with intelligent sensing and adaptive communication with physical and computational environments. They connect many different physical and informational subsystems that create the necessity for algorithms to quickly assess streamline analysis. Smart sensors could be linked together over wireless and cellular networks and carry larger volumes of data at reduced latency. Therefore, they are considered a key component for developing IIoT applications and providing efficient, reliable, and robust functionalities for a given system.

      Along with the increased capabilities of smart sensors, they have also become more flexible, power efficient, and miniaturized. Fusing sensing and local computing capabilities provides a solid framework for intelligent machines used in smart environments.

      1.8.1 Benefits of Smart Sensors in the Supply Chain

      Smart sensors change the way systems collect data. Aside from the aforementioned features, they offer three key benefits to Supply Chain 4.0:

       Operational efficiency: Smart sensors offer valuable added value to the system in real time, which enables the company to analyze and respond without human intervention. This results in operational efficiency through automation, improved demand planning, inventory control, asset management, and product life cycle management.

       Management and visibility: Rapid deployment of various smart sensors in IIoT enhances visibility across systems and assists in E2E supply chain management. This leads to cost reduction and generates incremental revenue. In addition, smart sensors connect end users more closely to the businesses and provide critical insight into customers’ behavior to enhance services. Another primary benefit offered by smart sensors within the Industry 4.0 framework is associated with the increased visibility of workflows and processes. The sensor measurements help real-time monitoring of equipment and assist in proactively receiving advance notices from potential problems or anomalies.

       Self-care and predictive maintenance (PM): Smart sensors could take advantage of artificial intelligence in Industry 4.0 and create self-identification, -diagnosis and -configuration sensors,