IoT Overview and Architecture
What the exam tests
The IoT Overview and Architecture domain checks that you can place devices, gateways, platforms, and applications into a coherent architecture and justify design choices. Expect questions on layered models, the device–edge–cloud continuum, and the difference between vertical and horizontal solutions.
Core rules
- The classic four-layer model is perception (sensors/actuators), network (access and transport), platform (device management, data, APIs), and application (business logic and UI).
- Edge sits between devices and cloud: it aggregates, filters, and can make local decisions when connectivity is poor.
- A horizontal IoT platform serves many industries; a vertical solution is built for one domain.
- Architecture choices trade off latency, bandwidth, power, cost, and security; no single layer does everything.
- Device identity and data models should be defined before scaling deployment.
Worked example
A utility wants smart electricity meters that report every 15 minutes and support remote disconnect. Map it: perception = meter with metering and relay; network = NB-IoT or PLC backhaul; platform = device registration, shadow, commands for disconnect, OTA; application = billing and outage dashboard. Reasoning: 15-minute reporting is low bandwidth, so LPWAN fits; remote disconnect is a command, so the platform must support reported/desired state; security needs device authentication and command authorization.
Traps
- Saying the gateway is the platform - gateways translate and pre-process; platforms manage devices and data at scale.
- Assuming every device is IP-capable; many use serial, Modbus, or ZigBee and need a gateway.
- Treating edge and cloud as interchangeable; edge is for local latency and resilience, cloud for global scale and analytics.