Evolution of Industrial Wireless Sensor Networks




Industrial automation powered by wireless sensor networks (WSN) is heralding the Industrial Internet of Things and Industry (IoT) 4.0. Key enabling cloud and wireless mesh networking technologies promise to bring multi-year battery life, IP addressability to machines and sensors, cloud-based provisioning and management systems, as well as fieldbus tunneling. Though currently prevalent wireless standards, such as WirelessHART and Wi-Fi will likely account for the majority of the industrial wireless network technologies, upcoming low power wide area (LPWAN) technologies are likely to grow in value for current and future applications.

FROM HARDWIRED CONNECTIONS TO SMART LINKS
Unlike datacenter communications and interoffice interconnections with patch cords and plenum Ethernet cables, the backbone for industrial-based communication require cables that are resistant to a number of environmental aggressors including chemicals, oils, moisture, vibration, and abrasion. This has been necessary for the fieldbus- and Ethernet-based hardwired industrial links but industrial wireless sensor networks (IWSN) are poised to replace much of that costly infrastructure with hundreds to thousands of modular sensor nodes.

There are a great variety of challenges to making this technology ubiquitous, for instance, the bit error rate (BER) of an industrial sensor node can stand between 10
-2 to 10-6 [1] while IEEE 802.3 Ethernet standards call for performance between 10-10 and 10-'2. For ultra-low latency time sensitive networking (TSN) applications such as factory automation with high speed robotic arms time synchronization protocols have to reliably achieve sub-millisecond cycles and sub-microsecond fitter for plant operations (e.g., PROFINET 10). WSN technology is limited by the amount of energy available to small battery-powered sensors, a limited bandwidth, and computational power; complex schemes that are traditionally used such as network time protocol (NTP) are not viable due to these restrictions [3].

Furthermore, many factory automation, process automation, and building automation facilities traditionally leverage the Purdue Enterprise Reference Architecture (PERA) for integrating applications in manufacturing operations and control as well as business systems. Each layer of hierarchy is highly custom and optimized independently to meet the requirements of specific tasks, this makes interoperability with industrial IoT (IloT) networks challenging thereby slowing the proliferation of this technology [4].

Still, the ability to add a level of abstraction between the hardware and software and the integration with internet protocol (IP) allow for scalable architectures to support a wide variety of industrial applications, this can allow for the economies of scale to arise much more rapidly than custom, proprietary systems. Additionally, the maintenance that comes with cables for interconnect is eliminated. In highly corrosive environments, the specially designed connector heads and cable jackets have to be inspected and maintained regularly to prevent network latencies and failures.



Figure 1 : Illustration of a WSN with a number of sensors that perform a diverse range of functions, a sink node that collects and sends data to the internet via cabling. nearly 8 of 10 users leverage wireless mesh topologies [2]. Source: http://article.sapub.org/10.5923.j.jwnc.20150501.03.html

AN EVOLUTION IN OPPORTUNITIES AND CHALLENGES
As shown in Figure 1, WSNs involve tiny wireless sensor nodes installed on industrial equipment to monitor its performance based on parameters such as vibration, temperature, proximity, power quality, and pressure. These nodes are often composed of a microcontroller, several sensors, communication modules, memory for data storage (e.g., EEPROM, SDcard), and a power source. Depending upon the access technology, the nodes can communicate to external systems using either Ethernet, Wi-Fi, ZigBee, Bluetooth, or GPRS/3G. Also known as the data acquisition (DAQ) sensor node as well as the gateway, the sink wirelessly receives the data and channels it through to the internet with an internal wired connection such as Ethernet. This prevents the necessity for each individual sensor node to store large amounts of data and provides a means for network backhaul in order to process and analyze data.

ENERGY CONSTRAINTS
Industrial WSNs face major resource constraints with energy, memory, and processing; most sensor nodes are battery powered limiting both the processing power of the sensor as well as the operational lifetime of the node.

There are many methods for energy harvesting that are being explored including energy generations through vibration, thermal energy, RF energy, and light. In outdoor environments, photovoltaic (PV) cells can be installed on the board for the primary source of energy where secondary energy storage is accomplished through a rechargeable battery bank and/or supercapacitors [5], [6].

On the plant floor, electric motors can account for more than 90% of energy output and also happen to be one of the biggest sources of energy on the plant floor to be exploited in terms of temperature and vibration. Thermoelectric energy can be harvested through the use of two dissimilar metals where a temperature gradient produces a current (Seebeck effect). Thermoelectric generators (TEG) composed of a number of n- and p-type semiconductor pellets use temperature gradients ranging from a few degrees to hundreds of degrees to generate energy. These temperature differentials can be from a human body or a machine to the ambient environment. Piezoelectric materials can convert vibrational and airflow strain into voltage. Ambient vibration can also be converted to power by means of magnetic induction with a magnet moving with respect to a coil.

Energy harvested through Microelectromechanical Systems (MEMS) or PV cells go through power conditioning in order to be stored in the secondary energy storage, through power management circuitry and finally to the load (sensor and radio). While there are many creative ways to autonomously generate energy to prevent the costly maintenance of replacing batteries, the unpredictability of the power source adds to challenge of the sensor node's reliability. Time sensitive information could be incorrectly transmitted or lost without the sensor node's replenishment of energy. Moreover, if a disproportionate amount of energy used up in transmission, sensors may not have enough power to detect the environment thereby degrading the potential determinism of a system [7].



Figure 2 : Most important features to WSN adopters based on ISA/ON World Survey done in 2012 and 2014 [2]

DETERMINISM AND RELIABILITY
A survey composed by the International Society of Automation (ISA) and ON World at the end of 2014 revealed that data reliability, security, and easy access to sensor data tend to the be
the most important features of a WSN (Figure 2). Interestingly enough, the concerns for battery life decreased while the need for IP addressability increased [2]. Two factors mainly contribute to the reliability of a WSN: mesh networking and channel hopping. Network clusters in which every node can communicate with multiple neighboring nodes inherently have a higher reliability through self-healing algorithms than linear, point-to-point topologies where if one node is rendered nonfunctional, the chain for transmission is broken.

Channel hopping is yet another failsafe for low power and lossy networks (LLNs) where nodes can use multiple channels within a given bandwidth in case there are transmission/reception challenges in select channels. This is defined in the media access control (MAC) layer within the IEEE 802.15.4e standard--the physical and data link defined basis for many WSN standards including WirelessHART, ZigBee, and ISA1 00.11 a.

Wired networks such as controller area networks (CAN) for automobiles would have to experience a BER of no more than 10-6 in order to have undetected corrupted messages occur less than once per year for the vehicle fleet [8] while the popular MIL-STD-1553B for avionics boosts BERs as low as 10-12. Most IWSN standards leverage a combination of Time Division Multiple Access (TDMA), Frequency Division Multiple Access (FDMA), and Carrier Sense Multiple Access with Collision Avoidance (CSMA/CA) of medium access control (MAC) protocols. The main difference between wired and wireless MAC protocols generally stems from the ability to detect collisions on the medium while sending (e.g., CSMA/ CD) [9]. Since this is not possible over the wireless medium, quality of service (QoS) analysis can be leveraged in IWSNs to measure packet loss, bandwidth, and delay. Moreover, additional MAC protocols can be utilized to increase the determinism of WSN.

Preliminary models are also being proposed that leverage binary countdown protocols [9], employ a collisionfree MAC protocol [10], attempt to approximate carrier sense multiple access with collision detection (CSMA/ CD) using the proposed carrier sense multiple access with collision notification (CSMA/CN) [11], uses an additional carrier sensing (ACS) algorithm to enhance the carrier sensing mechanism in the IEEE 802.15.4 CSMA/CA protocol [12], and leverages new channel access mechanism for a low latency deterministic network (LLDN) superframe in a star topology [13]. The CSMA/CA protocols generally suffer energy waste due to collisions and unpredictable end-to-end delays so TDMA mechanisms are employed in standards such as WirelessHART and ISAII 00.1 la for a more assured QoS with reservationbased medium access. Improvements over these standards are proposed that use a time-synchronized mesh network with short time slots where the device and overarching network operations are synchronized [14].

Table 1 : Dimensions of Security

SECURITY
Security ranks amongst the top concerns for IWSN end users (Figure 2). As shown in Table 1, there are several major aspects to security including data confidentiality, integrity, availability, freshness, and authenticity [15]. Strengthening all these aspects of security protect an IWSN against both
passive (e.g., transmission eavesdropping and sniffing) and active attacks (e.g., physical modification, Denial of Service, data falsification, and interruptions of service).

Figure 3 : Depiction of the various factions of Industrial WSN end users according to wireless systems standards and strategies [2]

WSN STANDARDS
As shown in Figure 3, as of 2014 one in four WSN adopters utilize the WireIessHART topology with the high 99% network reliability while one in ten are leveraging the ISA100.11a specification. However, in the past two years, ISA100.11 a adoption has increased 67% for its flexible time scheduling and software tunneling [16]. For low powered and long reach Low Power Wide Area Network (LPWAN) technology has growing interests. This topology boasts up to 1 0-year battery-powered wireless sensors with communication links up to 20 miles. While this technology may not be best-suited for secure, timesensitive and high reliability applications, it ranks highly in ease of use and scalability.

CONCLUSION
From environmental sensing, to condition monitoring, and process automation, IWSN service a broad range of applications. While ZigBee and MiWi generally service home automation applications, WirelessHART and ISA100.11 a are specifically designed for an industrial environment. Traditional wired industrial architectures do experience a greater level of determinism and a level of scalability with industrial Ethernet. Still, IWSNs surpass any wired network in modularity, ease of use, and cost-effectiveness.

REFERENCES
1. V. C. Gungor and G. P. Hancke, "Industrial Wireless Sensor Networks: Challenges, Design Principles, and Technical Approaches," in IEEE Transactions on Industrial Electronics, vol. 56, no. 10, pp. 4258-4265, Oct. 2009.

2. https://www.isa.org/ intech/201504web/

3. Z. Dengchang, A. Zhulin, and X. Yongjun, "Time Synchronization in Wireless Sensor Networks Using Max and Average Consensus Protocol", in International Journal of Distributed Sensor Networks Volume 2013.

4. https://www.plantservices.com/articles/2016/au-iiot-automation-zonesmart-device-ecosystem/

5. L. Vracar, A. Prijic, D. Nesic, S. Devic, and Z. Prijic, "Photovoltaic Energy Harvesting Wireless Sensor Node forTelemetry Applications Optimized for Low Illumination Levels", in Electronics 2016, 5, 26.

6. D. Antolin, N. Medrano, B. Calvo, and P. A. Martinez, "A Compact Energy Harvesting System for Outdoor Wireless Sensor Nodes Based on a Low-Cost In Situ Photovoltaic Panel Characterization-Modelling Unit", in Sensors 2017, 17, 1794.

7. L. Lei, Y. Kuang, X. S. Shen, K. Yang, J. Qiao and Z. Zhong, "Optimal Reliability in Energy Harvesting Industrial Wireless Sensor Networks," in IEEE Transactions on Wireless Commu
nications, vol. 15, no. 8, pp. 5399 5413, Aug. 2016.

8. Tran, Eushiuan. MS Thesis, "Multi-Bit Error Vulnerabilities in the Controller Area Network Protocol". Carnegie Mellon University, 1999. Web. 19 Oct. 2017.

9. Christmann, Dennis. Distributed Real-time Systems - Deterministic Protocols for Wireless Networks and Model-Driven Development. Diss. University of Kaiserslautern, 2015. Web. 19 Oct. 2017.

10. Nuno Pereira, Bjorn Andersson, and Eduardo Tovar. WiDom: A Dominance Protocol for Wireless Medium Access. IEEE Trans. Industrial Informatics, 3(2):120-130, 2007.

11. Sen, S.; Choudhury, R.R.; Nelakuditi, S. CSMA/CN: Carrier Sense Multiple Access With Collision Notification. IEEE/ACM Trans. Netw. 2012,20,544-556.

12. Lee, B.H.; Lai, R.L.; Wu, H.K.; Wong, C.M. Study on additional carrier sensing for IEEE 802.15.4 wireless sensor networks. Sensors 2010,10,6275-6289.

13. P. K. Sahoo, S. R. Pattanaik, and S. Wu, "Design and Analysis of a Low Latency Deterministic Network MAC for Wireless Sensor Networks", in Sensors 2017, 17, 2185.

14. M. Nixon, D. Chen, T. Blevins and A. K. Mok, "Meeting control performance over a wireless mesh network," 2008 IEEE International Conference on Automation Science and Engineering, Arlington, VA, 2008, pp. 540-547.

15. M. J. Jain, "Wireless Sensor Networks: Security Issues and Challenges", in IJCIT, Vol. 2, Issue 1, 2011.

16. https://www.researchandmarkets. com/research/748p7w/industrial


Article : Mark Miller, Wireless Product Manager, L-com Global Connectivity

November 2017



   


Home - Search - Suppliers - Links - New Products - Catalogues - Magazines
Problem Page - Applications - How they work - Tech Tips - Training - Events -