Wireless low-power IIoT sensor networks differentiated

Mesh networks and low-power WANs are for different IIoT use cases and installation settings.

By Albert Behr April 30, 2019

Driven by pervasive, battery-powered sensor networks that capture data at unprecedented granularity, the Industrial Internet of Things (IIoT) is reshaping the industrial connectivity landscape. From remote monitoring and energy management to worker safety and environmental sensing, IIoT applications demand wireless infrastructure that satisfies power consumption, scalability, mobility, and cost requirements without compromising carrier-grade reliability.

Low-power mesh networks based on the IEEE 802.15.4 standard and low power wide area networks (LPWANs) are the two leading edge connectivity options. While both cater to low-throughput IIoT applications, they vary greatly in network criteria. By understanding their differences, system designers can better decide which wireless option matches their architecture and use cases.

Mesh networks

IEEE 802.15.4 is a radio standard, defining physical and medium access control (MAC) layers in low-rate wireless personal area networks (LR-WPAN). Though the IEEE 802.15.4 standard allows for operations in different license-free industrial, scientific, and medical (ISM) bands, most solutions built on this standard are tuned for the 2.4 GHz band only. Examples include WirelessHART, ISA-100.11a, and Zigbee, all commonly adopted in industrial applications.

As the 2.4 GHz operations mode provides a very limited range of only 10 to 100 meters, these solutions resort to mesh topology to improve overall network footprint. Typically, a signal hops through multiple devices until it reaches the gateway. In full mesh networks like WirelessHART, all sensor nodes have the routing ability to relay data from other nodes. In partial mesh networks like Zigbee or ISA-100.11a, only selected nodes can act as routers.

Reliability in a mesh network is achieved through its self-healing capability. If a router or sensor node fails to function, the message can be re-routed through another path. To avoid package collisions and further improve robustness, approaches such as “listen before talk” or time-synchronized communications are employed.

On the downside, the 2.4 GHz band is highly congested with multiple sources of electromagnetic noise, including Wi-Fi hubs, Bluetooth devices, microwave energy, radio frequency (RF) lighting, industrial heaters and welding equipment. While WirelessHART and ISA-100.11a employ channel hopping or frequency hopping to improve interference immunity, this approach doesn’t help if all 2.4 GHz channels are saturated. Furthermore, the weak penetration capability of 2.4 GHz RF signals means that multipath propagation in structurally dense industrial settings can degrade range and signal quality.

By reducing the data rate to a maximum of 250 Kbit/s, IEEE 802.15.4 solutions aim to significantly lower power consumption. The relaying nature of mesh topology is, however, inherently energy-intensive. In full mesh networks, nodes placed close to the gateway or at strategically important positions carry substantial relaying traffic. They are highly vulnerable to battery failures, which risks disrupting major parts of the network.

Designing and managing mesh networks are a major undertaking. Additional devices must be installed as dedicated routers to attain the desired range, thereby adding redundancy and costs. High router density is required particularly in industrial settings with great physical barriers to offset the weak penetration and ensure network reliability. In manually configured mesh networks, users must calculate and configure different routing paths for each node, and manually handle any disruptions. The inflated complexity and costs due to redundant device density and connections hamper scalability and limit the applicability of these networks to medium-range use cases only.

Self-organizing solutions like WirelessHART simplify planning and operations but have their own scalability issues. Self-configured paths are not always the most optimal, which in turn, increases traffic and power consumption as more hops are created. On top of that, the congested 2.4 GHz airway, with its high level of interference, potentially constrains network capacity. As a result, in practice, the scalability of mesh solutions remains a few hundred devices per gateway, at best.

Lack of mobility support is another consideration for these networks. The IEEE 802.15.4 protocol is intended for communications from static devices only and are not applicable for IIoT use cases with mobile-end devices.

Low-power WANs

LPWANs are a family of technologies for low-throughput data communications over long range while consuming minimal energy. Though the data rates vary significantly among different solutions, they are usually lower than IEEE 802.15.4 networks. Except for cellular-based LPWANs, most technologies leverage the sub-GHz ISM bands, which are less crowded than the 2.4 GHz and enable better signal propagation.

Sub-GHz radio waves operating at narrow channel bandwidth experience less attenuation, bend farther around obstacles and provide better penetration through buildings. This increases signal performance in environments with multipath propagation caused by steel, metal, glass and other physical obstructions; and enables kilometers of range in LPWANs.

Thanks to their extensive range, LPWANs can be deployed in a one-hop star topology, which is far more power-efficient and easier to manage than the mesh topology. Since nodes do not have to constantly stay awake to relay messages, they can be kept in sleep mode much longer, thereby minimizing energy usage. Also, the lightweight MAC protocol and asynchronous communication drastically reduce overhead and power consumption per transmission.

Another major advantage of LPWANs is the low total cost-of-ownership. The simple waveforms minimize hardware design complexity and thus device costs, while the star topology combined with long range, reduce the requirement for expensive infrastructure (i.e., a base station). Ultra-low power consumption also reduces the need for battery replacement and associated maintenance costs.

Though outperforming IEEE 802.15.4 solutions in terms of range, battery life and costs, many LPWAN solutions, especially those using the license-free spectrum, cannot guarantee carrier-grade reliability. Specifically, long transmission (i.e., on-air) time resulting from slow data rates, coupled with asynchronous communication, increases the likelihood of message collisions and packet errors. Network performance quickly deteriorates as the number of devices and co-channel traffic escalate. Certain LPWAN technologies also suffer from low spectrum efficiency that constrains network capacity and scalability.

Telegram splitting ─ an emerging European Telecommunications Standards Institute (ETSI) standard for low throughput networks ─ promises to effectively overcome these challenges in license-free LPWANs. By reducing the on-air time and adopting robust techniques like frequency hopping and channel coding, telegram splitting provides a huge network capacity of millions of daily messages per gateway, regardless of the number of end nodes. At the same time, packet error rates are minimized in the presence of high radio interferences. What’s more, the technology offers excellent mobility support at up to 120 km/h.

Operating in the licensed spectrum, cellular LPWANs (e.g., NB-IoT, LTE-M) based on Third Generation Partnership Project (3GPP) standards are alternative reliable solutions. Nevertheless, it is worth noting quality-of-service and scalability in cellular LPWANs are achieved at the expense of comparatively higher power consumption and costs. Likewise, adopters are dependent entirely on operators’ network footprint, which is still far from ubiquitous.

Takeaway decisions

While IEEE 802.15.4-based mesh networks and LPWANs are both meant to support battery-powered sensor networks, each is suitable for different IIoT use cases and installations settings. The 802.15.4 solutions are a better fit in medium-range and medium-sized applications where nodes are mostly fixed and positioned in proximity with each other. The higher data rates and relatively low latency of these networks make them an ideal alternative to expensive wired networks in certain industrial automation and controls applications.

LPWANs are a better option for geographically dispersed campuses with challenging topography and greater physical obstructions, thanks to excellent range and penetration performance. They offer a more scalable, cost-effective and power-efficient solution for latency-tolerant IIoT use cases requiring only periodic data transmissions. Examples include condition monitoring, predictive maintenance, environmental sensing and energy management. With exceptional mobility support, certain LPWANs also offer viable connectivity for various worker safety and vehicle management use cases.

Though different from IEEE 802.15.4-based networks, not all LPWAN solutions are based on rigorous, worldwide proven industry standards. So far, there have been only two standardized LPWANs: one is cellular LPWANs implementing 3GPP standards; the other is MIOTY, a solution implementing the ETSI-standard telegram splitting technology (TS 103 357). Unlike proprietary protocols, solutions based on industry standards are verified for their quality of service and scalability while helping avoid the problem of vendor lock-in. Eventually, this helps companies better secure long-term return on investment (ROI) from their IIoT projects.

This article appears in the IIoT for Engineers supplement for Control Engineering and Plant Engineering. See other articles from the supplement below.

ONLINE extra

About the author

Albert Behr is founder and CEO of Behr Technologies, the worldwide licensee of MIOTYTM – the new commercial standard in LPWAN technology for Industrial IoT. With 30 years of professional experience, Albert has led the commercialization, financing, and operational execution of successful technology companies.

Original content can be found at Control Engineering.

Author Bio: Albert Behr is founder and CEO of Behr Technologies.