IoT Edge Computing and LiteOS
What the exam tests
Edge computing rationale and Huawei LiteOS components: kernel, drivers, components, and edge-cloud collaboration.
Core rules
- Edge computing runs compute near data sources to cut latency, save backhaul, and survive disconnection; it complements, not replaces, cloud.
- Edge nodes run local rules, protocol conversion, filtering, and AI inference; cloud handles global models, fleet management, and long-term storage.
- LiteOS is a lightweight real-time OS for MCUs: small footprint, fast boot, power management, and components such as sensors, connectivity, and OTA.
- LiteOS is not Linux; it targets constrained devices, so avoid assuming POSIX or full networking stacks.
Worked example
A production line needs to stop a motor within 50 ms when vibration exceeds a threshold. Put the rule on an edge node next to the PLC: local sensor read, threshold check, digital output. Reasoning: 50 ms is too tight for a round trip to cloud; the edge can act offline; the cloud still receives summaries for trend analytics and model updates.
Traps
- Saying edge replaces cloud; it offloads latency-critical work but cloud remains for scale.
- Assuming LiteOS is a general Linux distribution; it is an RTOS for MCUs.
- Forgetting edge security: physical access and stale firmware are real risks.