The value of an Internet of Things (IoT) device depends not only on its sensor, actuator, or embedded controller, but also on its ability to exchange information reliably. Connectivity technology determines how far a device can communicate, how much data it can transmit, how long its battery can last, how much infrastructure it needs, and how securely it can be operated.

No single technology is best for every smart electrical system. A protection relay in a substation may require deterministic, noise-resistant wired communication; a temperature sensor in a building may favor low-power wireless mesh networking; and a remote water meter may require wide-area cellular or LPWAN connectivity. Engineers therefore select communication technologies by matching technical requirements to environmental, economic, regulatory, and maintenance constraints.

This chapter compares Ethernet, Wi-Fi, Bluetooth Low Energy (BLE), Zigbee, Thread, LoRaWAN, NB-IoT, LTE-M, 5G, RS-485, CAN, and power-line communication (PLC). It also introduces the criteria used to evaluate them and applies those criteria to an illustrative smart-building design.

Learning Objectives

1. A Framework for Comparing Connectivity Technologies

A communication link transports information between nodes. In an IoT architecture, a node may be a sensor, actuator, gateway, controller, cloud service, or human-machine interface. The communication stack usually includes:

A technology name does not always identify a complete stack. Ethernet, for example, primarily describes local wired networking, while IP, TCP, UDP, and MQTT may operate above it. RS-485 specifies an electrical signaling method; it does not by itself define addressing, message meaning, or application behavior.

Range and coverage. Range is affected by transmit power, antenna gain, frequency, cable attenuation, obstacles, and receiver sensitivity. A stated nominal range is not a guarantee inside a steel building or industrial plant.

Bandwidth and latency. Bandwidth is the rate at which data can be transferred, whereas latency is the delay before data is delivered. A sensor transmitting a few bytes every minute needs little bandwidth but may still require bounded latency if it controls a safety function.

Topology. Common topologies are point-to-point, star, bus, tree, and mesh. Mesh networks can extend coverage and provide alternate routes, but routing increases complexity and energy use.

Reliability and determinism. Reliability concerns successful delivery over time. Determinism concerns whether delivery time can be bounded. Wireless contention and retransmissions can make latency variable; a properly engineered wired bus may provide more predictable behavior.

Power consumption. Battery-operated nodes favor short transmissions, sleep modes, low protocol overhead, and efficient network discovery. Mains-powered gateways can support more computationally demanding protocols.

Interference and propagation. Radio systems share spectrum with other devices and are affected by walls, machinery, multipath propagation, and electromagnetic interference (EMI). Wired systems avoid radio congestion but can suffer from ground loops, crosstalk, surge damage, and cable coupling.

Cost and deployment. Total cost includes modules, antennas, cabling, gateways, installation, commissioning, spectrum or subscription charges, cybersecurity maintenance, and eventual replacement.

where P_r is received power, P_t is transmitted power, G_t and G_r are transmit and receive antenna gains, L_p is propagation loss, and L_o represents other losses, all expressed in decibels or dBm consistently. Communication is feasible when received power exceeds the receiver sensitivity by an adequate fade margin. The equation is a planning model, not a guarantee of service; buildings, interference, antenna orientation, and time-varying fading must also be considered.

2. Wired Technologies

2.1 Ethernet

Ethernet is the dominant wired local-area networking technology. Modern Ethernet commonly uses twisted-pair copper or fiber optic cable and supports IP networking, industrial protocols, building automation, and high-speed backhaul. Power over Ethernet (PoE) can deliver electrical power and data through a suitable twisted-pair connection, simplifying installation of access points, cameras, controllers, and sensors.

Ethernet provides high bandwidth, low latency, mature diagnostics, and strong physical security when cables are protected. Switched Ethernet normally uses a star topology: devices connect to a switch, and the switch forwards frames. Fiber provides long distance, electrical isolation, and resistance to EMI, whereas copper is less expensive and easier to terminate over typical building distances.

Limitations include cable installation, connector and switch costs, vulnerability to physical cable damage, and the need for careful grounding and surge protection in industrial environments. Ordinary Ethernet is not automatically deterministic; time-sensitive industrial applications may require specialized mechanisms such as time-sensitive networking (TSN), careful traffic engineering, or a dedicated control network.

2.2 RS-485

RS-485 is a balanced differential electrical interface widely used in industrial instrumentation and building automation. It commonly supports a multidrop bus with a pair of signal conductors and can operate over substantial cable lengths at reduced data rates. A terminator is normally placed at each physical end of the bus, not at every device. Short stubs and appropriate biasing help prevent reflections and undefined bus states.

RS-485 is not a complete communication protocol. Modbus RTU is a common protocol carried over it, but other proprietary or standard protocols are possible. Its advantages are low interface cost, good noise rejection, simple wiring, and support for multiple nodes. Its limitations include shared-bus access, configuration sensitivity, relatively low bandwidth compared with Ethernet, and lack of inherent encryption or authentication.

2.3 Controller Area Network (CAN)

CAN was developed for robust distributed control, especially in vehicles, and is also used in machinery, energy systems, and embedded controls. Nodes broadcast frames identified by message identifiers. Arbitration allows a higher-priority message to gain access without corrupting the frame of a lower-priority message. Error detection, automatic retransmission, and fault confinement are important strengths.

CAN uses a terminated differential bus and is well suited to short, noisy environments with many controllers exchanging compact messages. CAN bus loading and identifier priorities must be engineered carefully. Classical CAN has limited payload per frame, while CAN FD permits larger payloads and higher data-phase rates. CAN does not inherently provide end-to-end encryption, so security must be supplied by system architecture or higher-layer mechanisms.

2.4 Power-Line Communication

PLC transmits information over conductors that also carry electrical power. It is attractive where installing a separate communications cable is expensive or impossible. Applications include smart metering, street lighting, home automation, and distribution-grid monitoring.

The power network is a difficult communication medium. Loads switch on and off, impedance varies with frequency, and motors, converters, and switching supplies produce noise. Coupling circuits, filtering, isolation, and regulatory compliance are essential. Narrowband PLC can provide robust utility-area coverage at modest data rates; broadband PLC provides greater throughput but can be more sensitive to channel conditions and electromagnetic compatibility constraints.

PLC deployment requires attention to electrical safety, isolation, conducted emissions, surge withstand, phase coupling, and the possibility that transformers or circuit topology will block or attenuate the signal.

3. Short-Range Wireless Technologies

3.1 Wi-Fi

Wi-Fi provides relatively high throughput over a local area and is widely supported by consumer and industrial devices. It commonly uses a star topology in which stations associate with an access point. Wi-Fi is suitable for cameras, gateways, firmware updates, dashboards, and devices that already have mains power.

Its strengths include high bandwidth, mature IP support, broad equipment availability, and integration with existing enterprise networks. Its weaknesses include comparatively high energy consumption, contention in crowded spectrum, variable latency, and reduced range through reinforced concrete or metal structures. Security should use current enterprise or personal protections, strong credentials, network segmentation, and controlled onboarding. A separate IoT network or virtual LAN is often preferable to placing low-trust devices directly on a corporate network.

3.2 Bluetooth Low Energy

BLE is designed for short-range, low-power communication. A sensor can advertise small packets and remain asleep most of the time; a phone, gateway, or controller can scan and connect when needed. BLE supports point-to-point links, broadcasting, and mesh profiles for selected applications.

BLE is useful for wearables, commissioning, occupancy sensors, asset identification, and configuration of electrical equipment. It has low module cost and is widely available in mobile devices. It is less suitable for large continuous data streams or long-range infrastructure unless gateways and appropriate physical-layer options are deployed. Pairing, key management, and protection against unauthorized commissioning require careful design.

3.3 Zigbee

Zigbee is a low-power wireless technology commonly used in sensing, lighting, and building automation. It generally operates in the 2.4 GHz band and supports star and mesh configurations. A coordinator establishes the network, routers forward traffic, and end devices may sleep to conserve energy.

Mesh routing can extend coverage and provide alternate paths around an obstruction. However, routers must remain powered, and network performance depends on radio placement, channel selection, joining policy, and vendor interoperability. Zigbee networks require a well-managed trust center and network keys. The 2.4 GHz band is shared with Wi-Fi and other devices, so channel planning is significant.

3.4 Thread

Thread is an IPv6-based, low-power mesh networking technology intended for connected devices. It uses IEEE 802.15.4 radios and supports sleepy end devices, routers, and border routers. Because Thread is IP-based, it can integrate naturally with broader IP networks, although an application layer such as Matter, CoAP, or another protocol is still required.

Thread offers self-healing mesh operation, low energy consumption, and avoidance of a proprietary network-layer design. A border router connects the Thread mesh to other IP networks. Engineers must still plan powered router locations, commissioning, security credentials, and coexistence in the 2.4 GHz band.

4. Low-Power Wide-Area and Cellular Technologies

4.1 LoRaWAN

LoRaWAN is a low-power wide-area networking system using a long-range, low-bit-rate physical layer and a star-of-stars architecture. End devices communicate with gateways, which forward packets to a network server. Gateways are not ordinary mesh routers; multiple gateways may receive the same uplink, improving coverage and enabling network-level selection.

LoRaWAN is appropriate for small, infrequent messages such as environmental measurements, metering, and status reports over city-scale or rural areas. Its advantages include long battery life, broad coverage, and the possibility of private or public networks. Its constraints include low throughput, duty-cycle or regulatory limits in some regions, variable latency, downlink limitations, and dependence on gateway coverage. It is not normally appropriate for continuous audio, video, or tight closed-loop control.

4.2 NB-IoT and LTE-M

NB-IoT and LTE-M are cellular IoT technologies operated through cellular infrastructure. NB-IoT emphasizes low power, deep coverage, and small data volumes. LTE-M supports greater mobility, higher throughput, and lower latency than NB-IoT, with better suitability for firmware updates and mobile assets.

Both technologies can reduce the need for an organization to build its own wide-area network. They provide operator-managed coverage and SIM or eSIM-based identity. Costs may include device modules, subscriptions, data charges, and dependence on an operator’s service life-cycle. Coverage inside basements and utility enclosures must be verified rather than assumed. Power-saving modes can greatly extend battery life, but they also increase message delay and complicate reachability for downlink commands.

4.3 5G

5G is a family of cellular technologies supporting enhanced mobile broadband, massive machine-type communication, and, in selected deployments, ultra-reliable low-latency communication. Its actual performance depends on frequency band, network configuration, spectrum, traffic, and service agreement.

5G is useful for high-capacity gateways, mobile robots, video inspection, distributed energy resources, and applications requiring managed wide-area mobility. Private 5G networks can offer local control and tailored policies but require spectrum access, radio planning, core-network functions, skilled operation, and substantial investment. Marketing claims about ultra-low latency should not be treated as universal guarantees; safety-critical control requires measured, engineered performance and often a local fallback.

5. Comparative Evaluation

The following qualitative comparison is more useful than treating any single range or data-rate figure as universal.

Technology	Typical scope	Topology	Bandwidth tendency	Power tendency	Principal strengths	Principal constraints
Ethernet	Room to campus	Star/switched	High	Mains	Speed, reliability, IP integration	Cabling and installation
RS-485	Room to plant section	Bus	Low to moderate	Very low interface power	Inexpensive, noise-resistant	Shared bus; protocol is external
CAN	Machine or vehicle	Bus	Low to moderate	Low	Arbitration and error handling	Small frames; bus loading
PLC	Building or electrical feeder	Shared medium	Low to high, technology-dependent	Usually mains-powered	Reuses power wiring	Noise, isolation, topology
Wi-Fi	Building or local site	Star	High	Moderate to high	Availability and throughput	Congestion and variable latency
BLE	Personal area/room	Star, broadcast, selected mesh	Low to moderate	Very low	Phones, low cost, low energy	Short range and gateway need
Zigbee	Building	Star/mesh	Low	Low	Mesh sensing and control	2.4 GHz coexistence
Thread	Building	IPv6 mesh	Low	Low	IP-based self-healing mesh	Border routers and commissioning
LoRaWAN	Campus to regional	Star-of-stars	Very low	Very low	Long range and battery life	Low payload and downlink limits
NB-IoT	Wide area	Cellular star	Low	Low	Operator coverage and deep reach	Subscription and operator dependence
LTE-M	Wide area/mobile	Cellular star	Moderate	Low to moderate	Mobility and greater throughput	Subscription and coverage
5G	Local to wide area	Cellular	High to very high	Moderate to high	Capacity, mobility, private networks	Cost and deployment complexity

The most appropriate technology is often hybrid. For example, battery sensors may use Thread or Zigbee to reach a mains-powered gateway, while the gateway uses Ethernet, Wi-Fi, LTE-M, or 5G for backhaul. A smart meter may use PLC within a distribution network and cellular communication to a utility server.

6. Design Calculations and Engineering Decisions

6.1 Data-rate estimation

where N is the number of devices, B is payload size in bits per report, f is the number of reports per device in the interval, and T is the interval duration in seconds. Protocol overhead, acknowledgements, retransmissions, guard times, and encryption overhead must be added for capacity planning.

Worked example. Two hundred sensors send a 32-byte payload every 60 seconds. The payload rate is:

This average is modest, but it does not prove that a low-rate network is adequate. If all sensors transmit simultaneously, the instantaneous load is much higher. Randomized reporting, scheduling, packet overhead, and required reliability must therefore be included.

6.2 Energy estimation

where I_i is current in amperes during operating state i, V is supply voltage in volts, t_i is time in that state per day, and E_sleep includes sleep energy. The result is in joules when time is in seconds. Real design must include regulator losses, battery temperature, self-discharge, ageing, retransmissions, and peak-current capability.

A protocol with a low average current may still be unsuitable if its transmission bursts cause voltage collapse or if network attachment takes too long. Battery life is therefore a system property, not merely a radio-module specification.

7. Security, Safety, and Governance

Connectivity expands the attack surface of an electrical system. Threats include eavesdropping, spoofed commands, replay, unauthorized joining, denial of service, malicious firmware, compromised gateways, and physical access to communication ports.

Good practice includes unique device identity, authenticated onboarding, encryption in transit, secure key storage, signed firmware, protected debug interfaces, network segmentation, logging, vulnerability management, and a documented update process. Encryption alone does not guarantee safety: an authenticated but incorrect command can still damage equipment.

Industrial and utility systems require defense in depth. Protective trips and essential control functions should not depend solely on an external cloud or an unbounded wireless link. Local control, fail-safe states, watchdogs, redundant paths, and manual override may be necessary. Engineers must also consider electromagnetic compatibility, radio licensing, privacy law, critical-infrastructure rules, electrical installation codes, and data retention requirements.

8. Illustrative Case Study: Smart University Engineering Building

Background

A university plans to monitor energy use, indoor air quality, occupancy, lighting, and distribution-board status in a four-storey engineering building. The building contains reinforced concrete, laboratories with variable-speed drives, and an existing Ethernet backbone.

Problem

The design team must connect approximately 200 battery-powered environmental sensors, 40 lighting controllers, 12 electrical panels, and several supervisory gateways. The system must support dashboards and energy analytics, while emergency protection remains locally controlled.

Analysis

Battery sensors transmit small readings every minute and must operate for several years. They do not need video bandwidth. Lighting controllers require responsive local communication and should continue operating if the internet fails. Panel data requires robust wired connections because electrical rooms contain substantial EMI. Gateways need reliable backhaul to the building management server.

Approach and solution

The team selects Thread mesh for battery sensors because it provides low-power IPv6 networking and can route around local obstructions. Mains-powered lighting controllers act as mesh routers, while border routers are installed on each floor. Ethernet connects border routers and supervisory servers.

RS-485 with an appropriate industrial protocol connects panel meters to local gateways. Short, correctly terminated bus segments are used, and communication cables are separated from high-current conductors. Ethernet is used for gateways and supervisory equipment; Wi-Fi is reserved for portable commissioning tools rather than essential control.

The system uses local schedules and fallback lighting behavior. Cloud access is used for long-term analytics but is not required for local operation. Device identities, network segmentation, signed updates, and gateway logging are included in the design.

Outcome

The hybrid design reduces battery consumption and avoids installing a separate cable to every sensor. Ethernet provides a stable backbone, while RS-485 suits electrical-panel instrumentation. The use of local control limits the consequences of internet or cloud failure.

Learning points

9. Contemporary Developments and Best Practices

IoT deployments increasingly use gateways capable of translating among fieldbus, IP, cellular, and low-power wireless networks. Edge computing reduces latency and limits the volume of data sent to the cloud. IPv6 adoption, standardized device models, software-defined cellular networks, private 5G, and energy-aware scheduling are expanding design options.

AI can assist with radio anomaly detection, predictive maintenance, adaptive reporting intervals, and interference classification. However, AI does not remove the need for deterministic protection logic, validated data, cybersecurity, or engineering accountability. An opaque model should not be allowed to issue unbounded commands to critical electrical equipment.

Recommended practice is to begin with a requirements matrix specifying payload, reporting interval, latency, reliability, coverage, battery life, environmental conditions, security level, installation constraints, and lifecycle cost. Prototype in the actual building, measure received signal and packet loss, test worst-case interference, document failure modes, and design for replacement of obsolete modules. Treat connectivity as infrastructure with a maintenance plan rather than as a one-time product choice.

Chapter Summary

Connectivity technology determines the coverage, capacity, energy use, reliability, security, and cost of an IoT system. Ethernet offers high-speed, mature wired networking; RS-485 and CAN provide economical and robust field communication; PLC reuses electrical wiring but must overcome noise and isolation challenges. Wi-Fi provides high local throughput, while BLE, Zigbee, and Thread support low-power local sensing and control. LoRaWAN offers long-range, low-data-rate communication, whereas NB-IoT and LTE-M use managed cellular networks for wide-area IoT. 5G provides high capacity and mobility but may require substantial infrastructure and careful performance validation.

Technology selection must consider topology, interference, power, latency, deterministic behavior, security, regulatory constraints, installation, and lifecycle cost. Hybrid architectures are often superior to single-technology designs. In electrical systems, safety-critical functions should retain local control and fail-safe behavior even when communication or cloud services fail.

Key Terms

BLE: Bluetooth Low Energy, a short-range wireless technology optimized for low power.

Determinism: Ability to provide bounded and predictable communication timing.

Ethernet: Wired local-area networking technology using copper or fiber media.

LoRaWAN: Low-power wide-area networking system for small, infrequent messages.

LTE-M: Cellular IoT technology supporting mobility and greater throughput than NB-IoT.

Mesh topology: Network topology in which nodes can forward traffic for one another.

NB-IoT: Narrowband cellular IoT technology emphasizing coverage and low power.

PLC: Power-line communication, transmission of data over electrical power conductors.

RS-485: Differential electrical interface commonly used for multidrop industrial buses.

Wi-Fi: Wireless local-area networking technology commonly using access points.

5G: Fifth-generation cellular networking technology supporting high capacity and diverse service classes.

Review Questions

Short Answer Questions

Long Answer Questions

Analytical / Application Questions

Multiple Choice Questions with Answers

Question 1

Which technology is primarily a differential electrical interface rather than a complete communication protocol?

Question 2

A battery sensor sends 32 bytes every 60 seconds. What is its payload rate, excluding overhead?

Question 3

Which technology is most associated with message arbitration and fault confinement in distributed control?

Question 4

Which technology generally provides the lowest data rate while emphasizing long range and very low power?

Question 5

Question 6

Question 7

B. It generally has higher energy consumption than low-power sensor technologies

Correct Answer: B. It generally has higher energy consumption than low-power sensor technologies

Question 8

Which technology reuses electrical power conductors as the communication medium?

Question 9

Which cellular IoT technology is generally better suited than NB-IoT to mobility and higher throughput?

Chapter 5 Connectivity and Communication Technologies

Connectivity and Communication Technologies

Introduction