🗺️ Smart Field Map
📡 Sensor Readings
Live🖥️ Fog Nodes
3 Online📊 Moisture Timeline
🔧 Actuator Control
Middleware Controlled🔔 Fog Alerts
0 ActiveNo alerts. System nominal.
Fog Computing Architecture – Smart Irrigation
Three-tier hierarchical architecture enabling real-time, low-latency decision-making at the edge.
⚡ Why Fog?
- Reduces cloud round-trip from 250ms → 12ms
- Local decision making without internet dependency
- 96% bandwidth reduction to cloud
- Offline-resilient irrigation control
🔌 Middleware Stack
- Protocol Adapter – MQTT ↔ HTTP ↔ CoAP
- Data Filter – Noise reduction & aggregation
- Event Bus – Publish/subscribe messaging
- Security Layer – TLS + token auth
🍕 Network Slices
- Slice 1 – Critical Control (<5ms)
- Slice 2 – Sensor Data (QoS high reliability)
- Slice 3 – Analytics (best-effort)
Network Slicing Demonstration
One physical infrastructure · Three virtual network slices with isolated QoS guarantees.
🔌 Shared Physical Infrastructure
Middleware Components
Middleware bridges IoT devices and fog/cloud services — handling protocols, filtering, routing, and security.
MQTT·CoAP·HTTP
Noise & Dedup
Pub/Sub Broker
Rules + ML
🔌 Protocol Adapter
Translates between heterogeneous IoT protocols. Normalizes all sensor data into a unified JSON schema before processing.
🔍 Data Filtering
Applies Kalman filtering and threshold-based deduplication to reduce noise and bandwidth usage by up to 60%.
📬 Event Bus (Pub/Sub)
Decouples producers (sensors) from consumers (actuators, cloud, analytics) using a publish-subscribe messaging pattern.
🧠 Decision Engine
Rule-based + ML hybrid engine makes irrigation decisions locally at the fog node without requiring cloud connectivity.
📨 Live Message Stream
Real-timeSystem Event Logs
| Time | Source | Type | Message | Latency |
|---|