Unit 3: IoT-Design Requirements - Subjective Questions

Message communication protocols are rules and standards that define how data is formatted, transmitted, received, and interpreted between connected IoT devices, gateways, servers, and applications.

Their importance includes:

• Enabling reliable communication between heterogeneous devices.
• Supporting data exchange over wired and wireless networks.
• Managing limited bandwidth, processing power, memory, and battery capacity.
• Providing mechanisms for addressing, authentication, error handling, and delivery confirmation.
• Allowing IoT applications to operate across different vendors and platforms.

Important selection factors include:

• Communication model: Request-response, publish-subscribe, or event-driven.
• Message size and bandwidth: Lightweight protocols are preferred for constrained networks.
• Power consumption: Battery-operated devices require energy-efficient communication.
• Reliability: The protocol should support acknowledgments, retransmissions, or quality-of-service levels when required.
• Latency: Real-time applications require fast message delivery.
• Security: Encryption, authentication, authorization, and integrity protection are essential.
• Scalability: The protocol should support a large number of devices and messages.

The publish-subscribe model is a messaging pattern in which devices do not communicate directly with one another. Instead, messages are distributed through a central or distributed broker.

• A publisher is a device or application that generates and sends data to a named topic.
• A subscriber is a device or application that registers interest in one or more topics.
• A broker receives messages from publishers and forwards them to all eligible subscribers.
• A topic is a logical channel used to classify messages, such as building/room1/temperature.

For example, a temperature sensor publishes a reading to the topic factory/machine1/temperature. A monitoring dashboard and an alert service subscribe to that topic. The broker delivers the reading to both subscribers without the sensor needing to know their addresses.

Advantages include:

• Loose coupling between devices.
• Easy addition or removal of subscribers.
• Efficient one-to-many communication.
• Support for asynchronous communication.
• Better scalability in large IoT deployments.

This model is commonly associated with lightweight protocols such as MQTT.

MQTT, CoAP, and HTTP serve different IoT communication requirements.

MQTT is suitable for devices that frequently send small messages to multiple applications. It provides different delivery guarantees through quality-of-service levels.

CoAP is designed for constrained devices and low-power or lossy networks. It resembles HTTP in using resources and methods such as GET, POST, PUT, and DELETE, but it uses UDP to reduce overhead.

HTTP is widely supported and integrates naturally with web applications and REST services. However, its headers and connection requirements may consume more bandwidth and energy.

The choice depends on latency, reliability, device limitations, network conditions, scalability, and integration requirements.

WebSocket is a communication protocol that provides a persistent, full-duplex communication channel between a client and a server over a single TCP connection.

Its working process is as follows:

1. The client begins with an HTTP-based handshake request.
2. The server accepts the upgrade request and changes the connection from HTTP to WebSocket.
3. A persistent connection is maintained between both endpoints.
4. The client and server can independently send messages at any time.
5. The connection is closed when either endpoint sends a close frame.

Important characteristics include:

• Full duplex: Data can flow in both directions simultaneously.
• Low latency: It avoids repeated HTTP request-response cycles.
• Persistent connection: The connection remains active for continuous communication.
• Event-driven operation: Data can be pushed immediately when an event occurs.
• Support for text and binary data: Suitable for telemetry and control messages.

IoT applications include real-time dashboards, remote equipment monitoring, industrial control, smart-home automation, live alerts, and device configuration. WebSocket is especially useful when the server must continuously push updated data to users or applications.

The three techniques differ in how they deliver updates from a server to a client.

• HTTP polling: The client sends requests at regular intervals to check whether new data is available. It is simple but may generate unnecessary traffic and introduce delays between polling requests.
• HTTP long polling: The client sends a request, and the server keeps it open until new data becomes available or a timeout occurs. The client then receives the response and submits another request. This reduces unnecessary responses but still requires repeated connections.
• WebSocket: The client and server establish one persistent connection. Either side can send data immediately whenever required.

For a real-time IoT dashboard, WebSocket generally provides the lowest latency and most efficient two-way communication, provided that the network and devices can maintain persistent TCP connections.

An IoT system must satisfy several design requirements to operate effectively and safely.

• Connectivity: Devices must communicate reliably with gateways, networks, and cloud services.
• Interoperability: Devices using different manufacturers, operating systems, and protocols should work together.
• Scalability: The system must support the addition of large numbers of devices without major performance degradation.
• Energy efficiency: Communication, sensing, and processing should minimize battery consumption.
• Real-time performance: Applications such as industrial control and healthcare monitoring may require low latency.
• Reliability and availability: The system should continue operating despite intermittent connectivity or component failures.
• Security and privacy: Device identity, communication, data, and user access must be protected.
• Manageability: Devices require remote provisioning, configuration, monitoring, firmware updates, and fault diagnosis.
• Data management: The system must collect, filter, store, process, and analyze large volumes of sensor data.
• Cost effectiveness: Hardware, communication, deployment, maintenance, and cloud costs must be considered.

These requirements must be balanced because improving one aspect, such as security or reliability, may increase processing, energy, or financial costs.

IoT deployments face technical, operational, and social challenges.

• Heterogeneity: Devices differ in hardware, operating systems, data formats, and communication protocols.
• Limited resources: Sensors may have restricted memory, processing capability, battery power, and bandwidth.
• Intermittent connectivity: Wireless links can suffer from interference, weak signals, congestion, and physical obstructions.
• Scalability: Addressing, authenticating, updating, and monitoring thousands or millions of devices is difficult.
• Security threats: Unprotected devices can be attacked, taken over, or used as entry points into larger networks.
• Privacy concerns: Sensors can collect personal, location, health, and behavioral information.
• Data volume: Large numbers of devices generate continuous streams that require efficient storage and analysis.
• Device management: Firmware updates, replacement, configuration, and fault detection are challenging after deployment.
• Interoperability: Proprietary technologies may prevent systems from exchanging data effectively.
• Environmental conditions: Temperature, moisture, vibration, dust, and physical access affect device reliability.
• Cost and maintenance: Installation and long-term operation may be expensive.

Successful IoT design addresses these challenges through standards, edge processing, secure lifecycle management, redundancy, efficient protocols, and suitable hardware selection.

IoT services collect data from physical environments, process it, and provide monitoring, control, automation, or decision support.

• Smart homes: Smart lighting, thermostats, security cameras, motion detection, energy management, and appliance automation.
• Healthcare: Wearable monitoring devices, remote patient monitoring, medication reminders, emergency alerts, and asset tracking in hospitals.
• Agriculture: Soil-moisture monitoring, automated irrigation, livestock tracking, weather observation, and crop-condition analysis.
• Industrial IoT: Predictive maintenance, machine monitoring, production automation, worker safety, inventory tracking, and quality control.
• Smart cities: Traffic monitoring, intelligent parking, waste management, air-quality monitoring, public lighting, and water distribution.
• Environmental monitoring: Measurement of pollution, temperature, humidity, floods, forest conditions, and natural hazards.
• Logistics: Fleet tracking, route optimization, cold-chain monitoring, and package-condition sensing.

A typical IoT service includes sensors, communication networks, data-processing platforms, user applications, and actuators. The service converts raw observations into useful actions, such as sending an alert when a machine temperature exceeds a safe threshold or automatically starting irrigation when soil moisture becomes low.

A sensor detects a physical, chemical, or biological condition and converts it into an electrical signal that can be processed by an IoT device.

Sensors may be classified as:

• Temperature sensors: Measure thermal conditions.
• Humidity sensors: Measure moisture in air or materials.
• Pressure sensors: Measure force per unit area.
• Motion and proximity sensors: Detect movement or nearby objects.
• Light sensors: Measure illumination.
• Gas and chemical sensors: Detect gases, smoke, or chemical concentrations.
• Biometric sensors: Measure parameters such as heart rate, blood oxygen, or body temperature.
• Location sensors: Use technologies such as GPS to determine position.

The conversion process generally includes:

1. A physical quantity changes the sensor's electrical property or output.
2. Signal-conditioning circuits amplify or filter the signal.
3. An analog-to-digital converter samples the signal.
4. A microcontroller processes, calibrates, and may compress the data.
5. The communication module transmits the data to a gateway, edge device, or cloud platform.

Sensor selection depends on range, accuracy, resolution, response time, calibration needs, power consumption, environmental tolerance, and cost.

An antenna is a device that converts electrical signals into electromagnetic waves for transmission and converts received electromagnetic waves back into electrical signals.

Important antenna parameters include:

• Operating frequency: The frequency range at which the antenna performs effectively, such as the frequency bands used by Wi-Fi, Bluetooth, Zigbee, LoRaWAN, or cellular networks.
• Bandwidth: The range of frequencies over which acceptable performance is achieved.
• Gain: The ability to focus transmitted or received energy in a particular direction. Higher gain can increase range but may reduce coverage in other directions.
• Radiation pattern: Describes how energy is distributed around the antenna.
• Polarization: Describes the orientation of the electric field. Transmitting and receiving antennas should generally have compatible polarization.
• Impedance: The opposition to alternating current, commonly designed near in radio systems.
• Efficiency: The proportion of input power converted into radiated power.
• Size and form factor: Important for small wearable, embedded, and battery-powered devices.

Antenna performance is affected by enclosure materials, nearby metal, orientation, body absorption, and installation height. Proper antenna selection improves coverage, reliability, energy efficiency, and communication range.

Antenna gain represents how effectively an antenna concentrates radiated energy in a particular direction. A higher-gain antenna can provide a stronger signal in its preferred directions, which may increase communication range or improve link reliability.

The received power can be approximated using the Friis transmission relationship:

where:

• is received power.
• is transmitted power.
• and are the transmitter and receiver antenna gains.
• is the wavelength.
• is the distance between the antennas.

A higher gain can allow the transmitter to achieve a required received power using lower transmit power, thereby reducing energy consumption. However, gain is usually achieved by concentrating energy rather than increasing total energy. Therefore, high-gain antennas may have narrow radiation patterns and can be unsuitable for mobile or randomly oriented sensors.

In IoT systems, the design must balance:

• Required range.
• Coverage direction.
• Device orientation.
• Antenna size.
• Transmit power.
• Battery life.
• Regulatory limits.

An omnidirectional antenna is often suitable for devices that can communicate in many directions, while a directional antenna may be preferable for fixed point-to-point links.

Cloud computing is the on-demand delivery of shared computing resources, including servers, storage, databases, networks, and software, over a network with flexible scaling and usage-based billing.

Cloud computing supports IoT systems by providing:

• Device connectivity: Cloud gateways and messaging services receive data from distributed devices.
• Scalable storage: Sensor data can be stored for historical analysis and auditing.
• Data processing: Cloud resources can process large data streams and run complex analytics.
• Machine learning: Cloud platforms can identify patterns, detect anomalies, and predict failures.
• Remote access: Users can monitor and control devices through web or mobile applications.
• Device management: Provisioning, authentication, configuration, firmware updates, and monitoring can be centralized.
• Integration: IoT data can be connected to enterprise systems, databases, and business applications.
• Reliability: Replication and distributed infrastructure can improve service availability.

A common architecture places sensors and actuators at the edge, communication networks in the middle, and cloud services at the backend. Edge computing may be added to reduce latency, bandwidth usage, and dependence on continuous cloud connectivity.

Cloud computing processes and stores IoT data in centralized or distributed data centers, while edge computing processes data close to the sensors or devices that generate it.

Edge processing is preferred when:

• The application requires immediate responses, such as industrial safety control.
• Network connectivity is unreliable.
• Sending all raw data to the cloud is expensive or impractical.
• Data privacy requires local processing.
• Devices generate high-volume data, such as video or vibration signals.
• Local operation must continue during cloud outages.

Cloud computing remains valuable for long-term storage, large-scale analytics, model training, centralized management, and combining data from many locations. Many modern IoT systems use a hybrid architecture in which urgent decisions occur at the edge and broader analysis occurs in the cloud.

The three major cloud service models provide different levels of abstraction and management responsibility.

Software as a Service (SaaS):

• Provides complete software applications over a network.
• The cloud provider manages the application, runtime, infrastructure, updates, and much of the security.
• IoT examples include web dashboards, fleet-management applications, energy-monitoring systems, and predictive-maintenance applications.

Platform as a Service (PaaS):

• Provides a development and deployment platform for building IoT applications.
• It commonly includes device registries, message brokers, databases, analytics tools, APIs, and application runtimes.
• Developers focus on application logic instead of managing servers and operating systems.

Infrastructure as a Service (IaaS):

• Provides virtual machines, storage, networks, and other basic computing resources.
• The customer manages operating systems, middleware, applications, and much of the configuration.
• It provides maximum flexibility but requires greater administration.

In general, SaaS offers the least control and lowest management effort, while IaaS offers the greatest control and highest management responsibility. PaaS provides a balance that is often appropriate for developing scalable IoT services.

SaaS, PaaS, and IaaS differ mainly in how much of the technology stack is managed by the cloud provider.

For IoT:

• SaaS is suitable for organizations that need ready-to-use monitoring or management applications.
• PaaS is suitable for developers building custom IoT services using device-management, messaging, and analytics features.
• IaaS is suitable when an organization needs specialized software, custom network configurations, or full control over virtual servers.

The choice should consider cost, technical expertise, customization, scalability, compliance, portability, and the required level of operational control.

IoT security issues arise because devices are widely distributed, often resource-constrained, and connected to networks that may contain valuable data and services.

Major issues include:

• Weak authentication: Default, shared, or easily guessed passwords allow unauthorized access.
• Poor authorization: Devices or users may receive more privileges than necessary.
• Insecure communication: Unencrypted traffic can be intercepted, modified, or replayed.
• Unpatched firmware: Vulnerabilities remain exploitable when devices cannot receive timely updates.
• Insecure interfaces: Web panels, APIs, mobile applications, and debugging ports may expose attack paths.
• Malware and botnets: Compromised devices can be controlled or used in distributed denial-of-service attacks.
• Data privacy violations: Sensors may expose location, health, behavior, or household information.
• Physical tampering: Attackers with physical access may extract keys, alter firmware, or replace components.
• Insecure cloud storage: Poorly configured databases or services may expose collected data.
• Denial of service: Network flooding or radio interference can make devices unavailable.
• Lack of secure lifecycle management: Devices may remain active after ownership changes or support ends.

Security must be considered during design, deployment, operation, maintenance, and decommissioning.

The main security objectives in IoT are as follows:

• Confidentiality: Only authorized entities should be able to read IoT data. Encryption protects sensor readings, credentials, and control commands from eavesdroppers.
• Integrity: Data and commands should not be altered without detection. Message authentication codes, digital signatures, and secure protocols help detect modification.
• Authentication: A device, user, or service must prove its identity before communication or access is allowed. Certificates, secure keys, and strong credentials can be used.
• Authorization: After identity verification, the system must determine what actions are permitted. Role-based or attribute-based access control can restrict operations.
• Availability: Devices and services must remain accessible when needed. Redundancy, rate limiting, fault recovery, and denial-of-service protection support availability.

For example, in a smart lock system, authentication verifies the identity of a mobile application, authorization determines whether it may unlock the door, confidentiality protects communication, integrity prevents command alteration, and availability ensures that legitimate users can access the lock when required.

These objectives must be balanced with IoT limitations such as low memory, limited processing power, and battery constraints.

A secure IoT communication architecture should protect devices, networks, services, and data throughout its lifecycle.

A suitable architecture includes:

1. Sensor and device layer: Each device receives a unique identity and secure credentials during manufacturing or provisioning. Debug interfaces should be disabled or protected.
2. Local network layer: Devices communicate using secure wireless configurations and network segmentation. Unnecessary services and ports should be disabled.
3. Gateway or edge layer: A gateway authenticates devices, validates messages, filters data, performs local analysis, and limits access between device networks and enterprise networks.
4. Transport layer: Data should be protected using encryption and endpoint authentication, such as TLS for TCP-based protocols or DTLS for UDP-based protocols.
5. Cloud platform layer: The platform should use secure device registries, access-control policies, encrypted storage, logging, monitoring, and key management.
6. Application layer: Applications should enforce user authentication, least-privilege authorization, input validation, secure APIs, and privacy controls.
7. Lifecycle management: Secure firmware updates, vulnerability monitoring, credential rotation, backup, and secure decommissioning must be supported.

This defense-in-depth approach prevents reliance on a single security mechanism and reduces the impact of a compromised device or network segment.

Encryption transforms readable data into an unintelligible form so that unauthorized parties cannot understand it. It is required both when data is transmitted and when it is stored.

• Encryption in transit protects messages moving between sensors, gateways, cloud services, and applications.
• Encryption at rest protects stored sensor data, credentials, logs, and configuration files.
• Symmetric encryption uses the same secret key for encryption and decryption. It is efficient and suitable for constrained devices.
• Asymmetric cryptography uses a public-private key pair. It supports identity verification, secure key exchange, and digital signatures but generally requires more computation.

Key management is often more difficult than selecting an encryption algorithm. A secure system should provide:

• Unique keys or credentials for each device.
• Secure key generation and protected storage.
• Safe provisioning before deployment.
• Key rotation and revocation.
• Recovery procedures for lost or compromised keys.
• Protection against unauthorized extraction.
• Secure certificate validation where certificates are used.

Using the same hard-coded key in every device is dangerous because compromise of one device may compromise the whole deployment. Effective encryption therefore requires secure identity management and a complete key lifecycle.

IoT devices remain deployed for long periods, so secure firmware updates are essential for correcting vulnerabilities, improving functionality, and responding to new threats.

A secure update process should include:

• Authenticity verification: The device verifies a digital signature before installing firmware.
• Integrity verification: A cryptographic hash confirms that the firmware was not modified or corrupted.
• Trusted source: Updates are obtained only from an authorized vendor or management service.
• Confidentiality when necessary: Firmware may be encrypted to prevent unauthorized analysis or copying.
• Rollback protection: Attackers should not be able to install an older vulnerable version.
• Failure recovery: A backup image or dual-partition design can allow recovery after an interrupted update.
• Access control: Only authorized administrators or services should initiate updates.
• Audit logging: Update attempts, versions, success, and failure should be recorded.

Lifecycle management also includes secure provisioning, inventory tracking, configuration, monitoring, credential rotation, vulnerability response, ownership transfer, and secure decommissioning. A device should be removed from trusted systems and have sensitive credentials erased when it reaches the end of its service life.