IoT Connectivity and Networking

Introduction

Connectivity is the foundation of the Internet of Things (IoT). An IoT device may sense temperature, identify an object, control a motor, or monitor a patient, but its practical value depends on how reliably and economically it communicates. Connectivity determines whether data travels a few centimetres or several kilometres, whether a battery lasts days or years, whether messages arrive in milliseconds or seconds, and whether a network can continue operating when devices move or obstacles appear.

IoT connectivity is not a single technology. It is a design choice involving physical media, radio spectrum, protocols, network topology, gateways, service providers, security controls, energy consumption, and cost. A warehouse asset tracker, a smart electricity meter, a wearable health sensor, and an industrial robot require different communication solutions.

This chapter compares important wired and wireless technologies: Ethernet, Wi-Fi, Bluetooth Low Energy (BLE), Zigbee, Thread, RFID, NFC, LoRaWAN, NB-IoT, and cellular networks. The comparison focuses on range, bandwidth, latency, energy use, reliability, topology, and cost, while also considering practical deployment and security.

Learning Objectives

By the end of this chapter, students should be able to:

1. Explain the role of connectivity in an IoT architecture.
2. Distinguish wired, short-range wireless, low-power wide-area, and cellular technologies.
3. Compare Ethernet, Wi-Fi, BLE, Zigbee, Thread, RFID, NFC, LoRaWAN, NB-IoT, and cellular networks using appropriate criteria.
4. Select a suitable connectivity technology for a given IoT application.
5. Explain network topologies, gateways, addressing, latency, bandwidth, and energy constraints.
6. Identify major security, regulatory, operational, and lifecycle concerns.

1. Connectivity as an IoT Design Decision

An IoT communication system commonly contains four elements:

• Endpoint or node: A sensor, actuator, tag, controller, or embedded device.
• Access network: The wired or wireless link connecting the endpoint to a gateway or network service.
• Gateway or edge device: A device that aggregates traffic, translates protocols, performs local processing, or connects a local network to the Internet.
• IoT platform and application: Cloud or enterprise software that stores, analyses, visualises, and acts on data.

A useful conceptual path is:

Sensor or actuator → local connectivity → gateway or access point → IP network → IoT platform → application or operator

Not every system contains all these elements. A Wi-Fi sensor may connect directly to an IP network, while a passive RFID tag may communicate only with a nearby reader. A LoRaWAN end device normally communicates with a gateway, which forwards packets to a network server.

Important design criteria

Range is the distance over which a link can operate acceptably. It depends on transmit power, antenna design, frequency, obstacles, interference, data rate, and regulatory limits. Indoor range is usually lower and less predictable than open-air range.

Bandwidth is the capacity of a communication channel, commonly expressed in bits per second. It indicates how much data can be transferred, but it does not by itself determine application performance. A high-bandwidth link may still have poor performance if latency or packet loss is high.

Latency is the time between transmission and useful reception. Control applications often need low and predictable latency, whereas a soil-moisture sensor may tolerate several minutes.

Energy consumption includes sleep, processing, transmission, reception, and sometimes network-maintenance energy. Battery-powered nodes often spend most of their time asleep. A communication technology that requires frequent listening or network handshakes can reduce battery life even when individual transmissions are short.

Reliability includes successful delivery, coverage, resistance to interference, predictable timing, redundancy, and recovery after failure. Reliability is broader than nominal signal range.

Topology describes how devices are arranged. Common forms include point-to-point, star, mesh, tree, and cellular topologies.

Cost includes modules, antennas, gateways, cabling, installation, spectrum or subscription charges, cloud services, maintenance, security updates, and eventual replacement.

2. Wired Connectivity: Ethernet

Ethernet is a family of wired local-area networking technologies widely used in buildings, factories, data centres, cameras, controllers, and gateways. It generally uses twisted-pair copper or fibre-optic cable and supports IP communication directly.

Ethernet's major strengths are high throughput, low and relatively predictable latency, strong physical security, and reliable power delivery when Power over Ethernet (PoE) is used. PoE allows compatible network equipment to provide electrical power and data over the same cable. This is useful for cameras, access points, industrial terminals, and building-control devices.

Ethernet usually uses a star topology, with devices connected to a switch. A switched Ethernet network can be extended through additional switches, although excessive complexity, loops, and poor configuration must be controlled. Fibre is suitable for long distances, electrical isolation, and environments with electromagnetic interference; copper is often cheaper and easier to install over shorter distances.

Ethernet is less suitable for mobile or physically dispersed battery devices because cables add installation cost and restrict movement. It also requires a wired infrastructure to every endpoint or to nearby equipment.

Ethernet characteristics

3. Wireless Personal and Local-Area Technologies

3.1 Wi-Fi

Wi-Fi is based on the IEEE 802.11 family and is designed for wireless local-area networking. It commonly connects sensors, appliances, cameras, controllers, laptops, and gateways to an IP network. Wi-Fi offers substantially higher throughput than most low-power IoT technologies and is widely available in homes, offices, campuses, and public locations.

Its main advantages are high data capacity, mature IP support, broad device availability, and the ability to carry software updates, images, audio, and video. Its disadvantages include relatively high energy consumption, shared radio-medium contention, interference, coverage limitations, and dependence on access-point power and configuration.

Wi-Fi normally uses a star topology: stations communicate through an access point. Modern Wi-Fi power-saving mechanisms can improve battery operation, but Wi-Fi is generally less suitable than BLE, Zigbee, or Thread for tiny sensors that transmit a few bytes infrequently.

Wi-Fi performance varies greatly with frequency band, channel width, modulation, distance, walls, competing networks, and access-point load. A high advertised physical-layer rate should not be interpreted as guaranteed application throughput.

3.2 Bluetooth Low Energy

Bluetooth Low Energy is designed for short-range, low-power communication. It is used in wearables, beacons, health devices, locks, keyboards, sensors, and commissioning processes. BLE devices commonly advertise small packets, establish connections when required, and return to a low-power state.

BLE traditionally uses a central-peripheral relationship: a central device, such as a phone or gateway, manages connections with peripheral devices. BLE also supports broadcast-style advertising and, through Bluetooth Mesh, many-to-many managed flooding for suitable building and control applications.

BLE is attractive because smartphones and tablets commonly include it, modules are inexpensive, and energy consumption can be very low when communication is intermittent. Its limitations include short practical range compared with LPWAN technologies, limited throughput for some profiles, and dependence on a phone or gateway for Internet connectivity in many deployments.

3.3 Zigbee

Zigbee is a low-power wireless technology widely used in lighting, home automation, building control, and sensor networks. It is associated with IEEE 802.15.4 physical and link-layer technology and commonly supports star and mesh arrangements.

In a Zigbee mesh, end devices may sleep, while routers forward traffic and a coordinator performs network-management functions. Mesh routing can extend coverage and provide alternate paths, but routers require continuous power or substantially more energy than sleeping end devices. Network planning must account for channel interference, device density, routing tables, and recovery after a router fails.

Zigbee is well suited to modest data volumes, low-power operation, and local networks. It normally requires a hub or gateway for Internet and application integration. Interoperability depends on profiles, product certification, and implementation quality.

3.4 Thread

Thread is an IPv6-based, low-power mesh networking technology built for connected homes and buildings. It uses IEEE 802.15.4 radio technology but differs from Zigbee in its IP-oriented architecture and network-management approach. Thread networks generally include sleepy end devices, routers, and one or more border routers. A border router connects the Thread network to other IP networks without requiring application-layer protocol translation in the same way as some proprietary systems.

Thread provides self-healing mesh operation, low-power endpoints, and direct compatibility with IP-based designs. It is useful for sensors, locks, lights, and building devices. Its deployment still requires careful commissioning, security-key management, radio planning, and suitable border-router infrastructure.

Comparison of short-range technologies

The values in this table are design tendencies, not guaranteed specifications. Building materials, antenna placement, country regulations, radio congestion, and application settings can change actual results.

4. Identification and Proximity Technologies

4.1 RFID

Radio Frequency Identification uses a reader and a tag to identify objects. Passive tags obtain operating energy from the reader's radio field and do not normally require a battery. Active or battery-assisted tags can provide greater range and additional sensing capability.

RFID is particularly useful when the main requirement is identification rather than continuous networking. Applications include inventory management, access control, library systems, logistics, livestock tracking, and automated checkout. A reader may identify multiple tags, depending on the RFID system and its anti-collision method.

RFID range varies substantially with frequency band, tag type, reader power, antenna design, orientation, and materials. Metal and liquids can affect performance. RFID tags usually do not provide the general-purpose IP connectivity, high throughput, or long-distance communication of Wi-Fi or cellular networks.

4.2 NFC

Near Field Communication is a very short-range technology derived from high-frequency RFID principles. It is commonly used for contactless payment, access cards, device pairing, configuration, and secure short-range exchanges. The close operating distance is a feature as well as a limitation: it reduces accidental communication and supports intentional tap-based interaction.

NFC can support reader/writer operation, card-emulation operation, and peer-to-peer functions in suitable implementations. A passive tag or card may draw energy from the reader field. Security is not guaranteed merely by physical proximity; applications still require authentication, encryption where appropriate, secure elements, and protection against relay or unauthorised-reading attacks.

5. Low-Power Wide-Area and Cellular Connectivity

5.1 LoRaWAN

LoRaWAN is a low-power wide-area networking protocol designed for long-range, low-data-rate IoT communication. It commonly uses a star-of-stars architecture: end devices transmit over a long-range radio link to one or more gateways, and gateways forward traffic to a network server. End devices do not normally relay one another's traffic.

LoRaWAN is valuable for smart metering, agriculture, environmental monitoring, asset tracking, and city infrastructure where messages are small and devices may be widely distributed. It can provide long battery life and broad coverage, especially when data is infrequent.

Its limitations include low throughput, duty-cycle or spectrum-management constraints in some regions, variable and sometimes significant latency, limited suitability for continuous streams, and dependence on gateway coverage. Downlink communication can be particularly constrained in battery-operated modes. Network design must also consider regional frequency plans and gateway backhaul.

LoRaWAN includes security concepts such as device and application session keys, but secure provisioning, key storage, server protection, and lifecycle management remain essential.

5.2 NB-IoT

Narrowband Internet of Things is a cellular IoT technology designed for low-power, wide-area deployments. It operates within licensed cellular spectrum and benefits from operator-managed infrastructure, authentication, wide-area coverage, and controlled radio operation. It is suited to smart meters, environmental sensors, parking systems, and other applications that exchange small amounts of data.

NB-IoT can offer better coverage inside buildings than many local wireless technologies and does not require an organisation to install its own gateways. However, devices require an operator subscription or managed service, modem and certification costs may be higher than for simple local radios, and latency and mobility behaviour depend on network configuration and power-saving features. Device sleep modes can reduce energy use but may delay downlink communication.

5.3 General cellular networks

Conventional cellular networks, including technologies such as 4G LTE and 5G, support wide-area mobility, high capacity, and direct connectivity through operator infrastructure. They are appropriate for connected vehicles, mobile equipment, video systems, industrial gateways, public safety, and applications requiring managed wide-area service.

Cellular systems can provide strong authentication, mobility management, quality-of-service mechanisms, and broad geographic coverage. Their disadvantages may include subscription charges, modem and antenna costs, network dependence, energy consumption, and changing operator support over the device's lifetime. Cellular IoT design must distinguish a high-bandwidth mobile broadband service from specialised low-power cellular modes.

6. Comparative Decision Framework

No technology is universally best. A selection should begin with application requirements rather than technology preference.

A practical selection sequence

1. Define the payload and message frequency. A temperature value every hour is fundamentally different from video or closed-loop motor control.
2. Define the physical environment. Consider indoor walls, underground locations, metal structures, vehicles, weather, and electromagnetic interference.
3. Specify latency and reliability. Determine whether occasional delay is acceptable and whether the system must continue during network outages.
4. Estimate energy and maintenance constraints. Identify whether mains power, replaceable batteries, energy harvesting, or multi-year operation is required.
5. Determine ownership and coverage. Decide whether the organisation can install gateways, switches, and access points or prefers an operator-managed network.
6. Assess security and lifecycle requirements. Include identity, encryption, key rotation, firmware updates, certificate expiry, and technology availability.
7. Calculate total cost of ownership. Include installation, subscriptions, gateways, batteries, site visits, monitoring, and disposal—not merely module price.

7. Reliability, Latency, and Energy in Practice

A communication link is a system, not just a radio. Reliability may be improved through antenna placement, diversity, acknowledgement, retries, buffering, redundant gateways, alternate routes, local control, and monitoring. Retries improve delivery probability but consume energy and increase latency. Aggressive retransmission is therefore not automatically a good design.

For a message that must be delivered successfully, a system may use acknowledgements and a bounded retry policy. If the probability that one independent attempt succeeds is p, then the probability that at least one of n attempts succeeds is:

Here, P_success is dimensionless, p is the success probability of one attempt, and n is a positive integer. The independence assumption is important; interference bursts and coverage gaps can make attempts correlated, so the formula may overestimate real reliability.

For example, if one attempt succeeds with probability 0.8 and three attempts are statistically independent:

P_success = 1 − (1 − 0.8)³ = 1 − 0.2³ = 0.992

The theoretical probability is 99.2%, but practical engineering must also consider duplicate messages, battery cost, queueing, correlated interference, and acknowledgement failure.

Energy use is influenced by radio-on time. A simple conceptual energy relationship is:

where E is energy in joules, P is power in watts, and t is active time in seconds. This relationship applies when power is approximately constant during the interval. Real devices should account separately for sleep, transmit, receive, processing, and leakage energy. A low transmit power does not guarantee low total energy if the device remains awake for long periods.

8. Security and Governance

IoT networks enlarge the attack surface because devices may be numerous, physically accessible, inexpensive, and difficult to update. Security should be designed across the device, link, gateway, network, cloud, and application layers.

Important controls include:

• unique device identities rather than shared default passwords;
• secure commissioning and authenticated onboarding;
• encryption in transit and, where necessary, protection at rest;
• secure boot and signed firmware updates;
• least-privilege access for devices and services;
• network segmentation for operational technology;
• logging, anomaly detection, and incident response;
• key rotation, certificate management, and secure decommissioning;
• privacy minimisation for location, health, behavioural, and identity data.

Wireless spectrum is regulated nationally and internationally. Designers must observe permitted frequencies, transmit-power limits, duty-cycle rules, approved equipment, and electromagnetic compatibility requirements. Cellular and NB-IoT deployments also depend on operator contracts, regional coverage, subscriber identity management, and service continuity.

RFID, NFC, BLE beacons, and cellular trackers can reveal identity or location. Organisations should define retention periods, access permissions, consent processes where applicable, and transparent user communication. Safety-critical systems require fail-safe behaviour and should not rely solely on a cloud connection for immediate control.

9. Applied Case: Environmental Monitoring in a Technical College

Background

This illustrative case concerns a technical college wishing to monitor temperature, humidity, energy use, and water leakage across classrooms, laboratories, and storage areas. Some buildings have reliable network cabling; others are old structures with thick walls. The college wants battery-powered sensors to operate for several years and requires alarms for leaks and unsafe temperatures.

Problem

A single technology does not satisfy all requirements. Ethernet provides dependable connectivity but would require extensive cabling. Wi-Fi is already available, but many battery sensors would need frequent maintenance. LoRaWAN offers broad coverage, but its low data rate and variable downlink make it unsuitable for every control function. The college also needs secure access and local operation during Internet failures.

Analysis

The monitoring payload is small: sensor readings may be sent every five or fifteen minutes, with immediate event messages for leaks. Cameras are outside the sensor scope and require much greater bandwidth. Laboratory controllers may require low latency and predictable local communication. The buildings contain metal equipment that may attenuate radio signals.

Approach and solution

The proposed hybrid design is:

• Ethernet and PoE for gateways, network switches, important controllers, and selected fixed devices.
• Thread or Zigbee for indoor temperature, humidity, and occupancy sensors where a suitable mesh can be installed and powered routers are available.
• BLE for commissioning, maintenance, and portable instruments used by technicians.
• LoRaWAN for remote storage buildings and outdoor water-level sensors, with gateways connected through Ethernet or cellular backhaul.
• Wi-Fi for mains-powered devices that require higher throughput or already have certified Wi-Fi support.
• Local edge rules so a water-leak alarm can activate a valve or siren without waiting for a cloud response.

Every device receives a unique identity. Gateways are placed after a radio survey rather than simply at convenient electrical points. Firmware updates are tested before staged deployment. Sensor readings are buffered during backhaul outages, and alarms are prioritised over routine telemetry.

Outcome

The hybrid design reduces cabling while avoiding dependence on a single radio technology. Battery sensors use low-power communication, high-bandwidth devices use Wi-Fi or Ethernet, and remote low-data sensors use LoRaWAN. The college accepts the additional complexity of operating several technologies because the arrangement better matches the physical and operational requirements.

Learning points

1. Connectivity selection should begin with payload, latency, power, and environment.
2. A gateway can combine technologies without forcing every device onto one network.
3. Local control improves safety and resilience.
4. Total cost includes batteries, surveys, gateways, subscriptions, and maintenance.
5. Security and lifecycle management must be included before deployment, not added afterward.

10. Contemporary Developments and Best Practices

IoT networking is increasingly designed around IP interoperability, edge computing, device-management standards, and automated security. Thread and other IPv6-based technologies support integration across application ecosystems. Cellular networks continue to develop specialised low-power and high-reliability services. Wi-Fi evolution improves efficiency for dense and battery-aware deployments, although actual benefits depend on device and access-point support.

Artificial intelligence can assist network operations by detecting unusual traffic, predicting battery depletion, identifying failing links, and optimising maintenance schedules. AI does not replace sound radio planning or security controls. Training data may be incomplete, false alarms may disrupt operations, and automated decisions require human oversight in safety-sensitive environments.

Best practices include conducting a site survey, documenting assumptions, testing worst-case conditions, measuring real battery life, planning gateway capacity, separating control from monitoring traffic, and designing for replacement. Pilot deployments should include failure tests: loss of power, gateway outage, interference, expired credentials, unavailable cloud service, and corrupted firmware update.

Chapter Summary

IoT connectivity determines the achievable range, data capacity, latency, energy life, reliability, topology, security, and cost of an IoT system. Ethernet provides dependable, high-bandwidth wired communication and can deliver power through PoE, but installation restricts mobility. Wi-Fi offers high throughput and broad availability, with greater energy and contention concerns. BLE is efficient for short-range personal devices; Zigbee and Thread support low-power mesh networks, with Thread providing an IP-oriented architecture.

RFID and NFC are primarily identification and proximity technologies rather than general-purpose wide-area networks. LoRaWAN supports long-range, low-data-rate communication through gateways and is effective for dispersed, battery-powered sensors. NB-IoT and other cellular technologies provide operator-managed wide-area coverage, authentication, and mobility options, but introduce subscriptions and network dependency.

Technology selection should consider the complete application: payload, frequency, range, latency, reliability, energy, topology, environment, security, regulation, ownership, and lifecycle cost. Hybrid architectures are often superior to single-technology designs. Reliable IoT systems also use appropriate retry policies, buffering, local control, secure onboarding, firmware management, monitoring, and planned decommissioning.

Key Terms

Bandwidth: The data-carrying capacity of a communication channel.

BLE: Bluetooth Low Energy, a short-range technology designed for low-power devices.

Border router: A device that connects a Thread network to another IP network.

Cellular network: An operator-managed radio network divided into geographic cells.

Gateway: A device that aggregates traffic or connects different networks and protocols.

Latency: The time between sending data and receiving or using it.

LoRaWAN: A low-power wide-area networking protocol using gateways and long-range radio links.

Mesh topology: A network arrangement in which nodes can provide multiple communication paths.

NB-IoT: Narrowband Internet of Things, a cellular technology for low-power wide-area applications.

NFC: Near Field Communication, a very short-range technology used for proximity interaction.

PoE: Power over Ethernet, which carries electrical power and data over suitable Ethernet cabling.

RFID: Radio Frequency Identification using readers and electronic tags.

Reliability: The ability of a system to deliver required service consistently under stated conditions.

Star topology: A network in which devices communicate through a central node such as a switch or access point.

Thread: An IPv6-based low-power mesh networking technology.

Topology: The logical or physical arrangement of devices and links.

Total cost of ownership: All acquisition, installation, operation, maintenance, subscription, and retirement costs.

Zigbee: A low-power wireless technology commonly used in sensor and automation networks.

Review Questions

Short Answer Questions

1. What are the main components of a typical IoT communication path?
2. Why is Wi-Fi generally less suitable than BLE for many tiny battery-powered sensors?
3. Distinguish RFID from NFC in terms of range and common application.
4. What is the function of a gateway in an IoT network?
5. State two advantages and two limitations of LoRaWAN.

Long Answer Questions

1. Compare Ethernet, Wi-Fi, BLE, Zigbee, and Thread with respect to bandwidth, energy, topology, and typical applications.
2. Explain how range, latency, reliability, and energy requirements influence IoT connectivity selection.
3. Describe the architecture and practical uses of LoRaWAN, including the role of end devices, gateways, and network servers.
4. Compare NB-IoT with general cellular connectivity for wide-area IoT deployments.
5. Discuss the security and governance controls required for a professional IoT network.

Analytical / Application Questions

1. A farm needs soil sensors that transmit a small reading every 30 minutes over several kilometres, with battery life measured in years. Which technologies would you shortlist, and why?
2. A factory requires low-latency control of fixed motors, video monitoring, and wireless condition sensors. Propose a hybrid connectivity architecture and justify each technology.
3. A campus has thick concrete walls, existing Wi-Fi, several remote buildings, and a requirement for local operation during Internet outages. Design a connectivity strategy and identify its principal risks.

Multiple Choice Questions with Answers

Question 1

Which technology normally provides the highest and most predictable bandwidth and latency among the following?

A. Ethernet

B. NFC

C. LoRaWAN

D. RFID

Correct Answer: A. Ethernet

Question 2

Which technology is primarily designed for very short-range proximity interaction such as tap-based access or payment?

A. NB-IoT

B. NFC

C. LoRaWAN

D. Ethernet

Correct Answer: B. NFC

Question 3

What is a major characteristic of Thread?

A. It is an IPv6-based low-power mesh technology

B. It is a passive identification technology only

C. It is designed mainly for high-definition video

D. It requires a wired connection to every endpoint

Correct Answer: A. It is an IPv6-based low-power mesh technology

Question 4

In a typical LoRaWAN star-of-stars architecture, what forwards end-device traffic toward the network server?

A. A passive RFID tag

B. A LoRaWAN gateway

C. An NFC card

D. A BLE keyboard

Correct Answer: B. A LoRaWAN gateway

Question 5

If one independent transmission attempt succeeds with probability 0.8, what is the probability that at least one of three attempts succeeds?

A. 0.512

B. 0.800

C. 0.960

D. 0.992

Correct Answer: D. 0.992

Question 6

Which technology is generally most suitable for a mains-powered IoT camera requiring high data throughput?

A. Wi-Fi

B. NFC

C. RFID

D. LoRaWAN

Correct Answer: A. Wi-Fi

Question 7

What is the principal advantage of a passive RFID tag?

A. It can stream video over cellular networks

B. It normally obtains operating energy from the reader field

C. It always provides multi-kilometre IP connectivity

D. It requires a large rechargeable battery

Correct Answer: B. It normally obtains operating energy from the reader field

Question 8

Which factor is especially important when selecting NB-IoT for a deployment?

A. Operator coverage and subscription arrangements

B. The need for a passive tag antenna only

C. The absence of any network infrastructure

D. A requirement for centimetre-range communication

Correct Answer: A. Operator coverage and subscription arrangements

Question 9

In a mesh network, what is a key benefit of multiple paths?

A. They can provide alternate routes when a node or link fails

B. They eliminate all energy consumption

C. They guarantee unlimited bandwidth

D. They remove the need for security

Correct Answer: A. They can provide alternate routes when a node or link fails

Question 10

Which cost is part of total cost of ownership for an IoT connectivity solution?

A. Only the initial radio module price

B. Only the purchase price of sensors

C. Installation, subscriptions, maintenance, batteries, and retirement

D. Only the cost of application software

Correct Answer: C. Installation, subscriptions, maintenance, batteries, and retirement