Edge Nodes Reference — Hardware & Software Patterns for Smart City Sensor Networks

MQTT Protocol Support

• Versions: MQTT 3.1.1, MQTT 5.0
• QoS Levels: 0 (At most once), 1 (At least once), 2 (Exactly once)
• Security: TLS 1.2/1.3, certificate-based authentication, username/password
• Features: Persistent sessions, retained messages, last will and testament, message expiry, shared subscriptions (MQTT 5.0)
• Topic Structure: Hierarchical with wildcards (+, #) support
• Extensions: MQTT-SN for constrained devices

MQTT Configuration Example

# MQTT Broker Configuration on Edge Node

listener 1883
  bind_interface eth0
  max_connections 1000
  
listener 8883
  bind_interface eth0
  max_connections 1000
  tls_version tlsv1.2 tlsv1.3
  cafile /etc/mosquitto/ca.crt
  certfile /etc/mosquitto/server.crt
  keyfile /etc/mosquitto/server.key
  require_certificate true
  
persistence true
persistence_location /var/lib/mosquitto/
persistence_file mosquitto.db

log_dest file /var/log/mosquitto/mosquitto.log
log_type all

connection_messages true
log_timestamp true

allow_anonymous false
password_file /etc/mosquitto/passwd

CoAP Protocol Support

• Versions: CoAP RFC 7252, CoAP over TCP, CoAP over WebSockets
• Security: DTLS 1.2, Object Security for Constrained RESTful Environments (OSCORE)
• Features: Confirmable/Non-confirmable messages, resource observation, block-wise transfers
• Content Formats: JSON, CBOR, EXI, Plain text
• Extensions: CoRE Resource Directory, CoRE Link Format
• Interoperability: CoAP-HTTP proxying

CoAP Gateway Configuration Example

# CoAP Gateway Configuration on Edge Node

# General Settings
host: 0.0.0.0
port: 5683
secure_port: 5684
max_connections: 500

# DTLS Configuration
dtls:
  enabled: true
  psk_identity: "edge-node-identity"
  psk_key_file: "/etc/coap/psk.key"
  certificate: "/etc/coap/server.crt"
  private_key: "/etc/coap/server.key"
  ca_certificate: "/etc/coap/ca.crt"
  
# Resource Directory
resource_directory:
  enabled: true
  registration_lifetime: 86400
  
# HTTP Proxy
http_proxy:
  enabled: true
  port: 8080
  
# Logging
log_level: info
log_file: "/var/log/coap/coap-gateway.log"

Translation Components

• Message Format Conversion: Transforming between binary protocols, JSON, CBOR, and platform-specific formats
• Semantic Mapping: Normalizing different data models into the NeuraAtlas canonical data model
• QoS Translation: Mapping between different quality of service mechanisms
• Security Translation: Converting between security schemes while maintaining appropriate protection levels
• Metadata Enrichment: Adding contextual information like timestamps, location, and device identifiers

Translation Pipeline

1. Protocol Decoding: Parsing incoming messages according to source protocol specifications
2. Message Validation: Verifying message integrity and format correctness
3. Schema Transformation: Converting to canonical data model
4. Enrichment: Adding metadata and contextual information
5. Protocol Encoding: Formatting for target protocol
6. Delivery: Transmitting to destination with appropriate QoS

Filtering Techniques

• Threshold-based Filtering: Only transmitting values that exceed defined thresholds or change by significant amounts
• Statistical Filtering: Using statistical methods to identify and filter outliers
• Temporal Filtering: Reducing data rate through downsampling while preserving important patterns
• Semantic Filtering: Filtering based on data meaning and context
• Quality-based Filtering: Dropping low-quality readings based on sensor confidence metrics

Filter Configuration Example

# Data Filtering Rules

filters:
  - name: "temperature_threshold"
    sensor_type: "temperature"
    conditions:
      - type: "change"
        min_delta: 0.5
        time_window: 60
      - type: "threshold"
        min: -30
        max: 60
    action: "forward"
  
  - name: "traffic_downsampling"
    sensor_type: "traffic_counter"
    conditions:
      - type: "temporal"
        interval: 300
        aggregation: "sum"
    action: "aggregate_and_forward"
  
  - name: "noise_outlier"
    sensor_type: "noise_level"
    conditions:
      - type: "statistical"
        method: "zscore"
        window: 20
        threshold: 3.0
    action: "drop"

Aggregation Methods

• Temporal Aggregation: Combining readings over time windows (e.g., 5-minute averages)
• Spatial Aggregation: Combining readings from sensors in proximity
• Multi-sensor Fusion: Creating derived metrics from multiple sensor types
• Hierarchical Aggregation: Progressive summarization across edge node tiers

Aggregation Functions

• Statistical: min, max, mean, median, percentiles, standard deviation
• Counting: count, distinct count, frequency
• Cumulative: sum, running total, integration
• Rate-based: velocity, acceleration, derivatives

Aggregation Configuration Example

# Aggregation Pipeline Configuration

aggregations:
  - name: "district_temperature"
    inputs:
      - topic: "sensors/+/temperature"
        filter: "location.district = 'central'"
    window:
      type: "tumbling"
      size: "5m"
    operations:
      - type: "avg"
        output_field: "avg_temp"
      - type: "min"
        output_field: "min_temp"
      - type: "max"
        output_field: "max_temp"
      - type: "count"
        output_field: "sensor_count"
    output:
      topic: "aggregated/central/temperature"
      format: "json"
      
  - name: "traffic_junction"
    inputs:
      - topic: "sensors/junction42/+/traffic"
    window:
      type: "sliding"
      size: "15m"
      slide: "5m"
    operations:
      - type: "sum"
        field: "vehicle_count"
        output_field: "total_vehicles"
      - type: "avg"
        field: "speed"
        output_field: "avg_speed"
      - type: "custom"
        expression: "total_vehicles / (avg_speed + 0.1)"
        output_field: "congestion_index"
    output:
      topic: "analytics/junction42/traffic_summary"
      format: "json"

Event Detection Approaches

• Rule-based Detection: Using predefined conditions to identify events
• Pattern Matching: Identifying specific sequences or combinations of readings
• Anomaly Detection: Identifying deviations from normal behavior
• Trend Analysis: Detecting significant changes in trends
• Complex Event Processing: Correlating multiple event streams

Event Response Actions

• Immediate Notification: High-priority alerting for critical events
• Local Actuation: Triggering local responses without central system involvement
• Enhanced Monitoring: Increasing data collection frequency for areas of interest
• Event Logging: Recording detailed information about detected events
• Escalation: Multi-stage notification based on event persistence or severity

Event Detection Configuration Example

# Event Detection Rules

events:
  - name: "traffic_congestion"
    description: "Detect traffic congestion at intersection"
    condition:
      type: "complex"
      expression: |
        vehicle_count > 50 AND
        avg_speed < 10 AND
        duration("avg_speed < 15") > 300
    actions:
      - type: "notification"
        channel: "traffic_control"
        priority: "high"
        template: |
          Traffic congestion detected at {{location.name}}
          Vehicle count: {{vehicle_count}}
          Average speed: {{avg_speed}} km/h
          Duration: {{duration}} seconds
      - type: "actuation"
        target: "traffic_signals"
        command: "CONGESTION_PROGRAM"
        parameters:
          program_id: 3
          duration: 1800
    throttling:
      min_interval: 600
      
  - name: "air_quality_alert"
    description: "Detect dangerous air quality levels"
    condition:
      type: "threshold"
      field: "pm25"
      operator: ">"
      value: 150
      duration: 900
    actions:
      - type: "notification"
        channel: "environmental"
        priority: "critical"
      - type: "data_collection"
        action: "increase_frequency"
        targets: [
          "air_quality_sensors",
          "weather_stations"
        ]
        parameters:
          frequency: 60
          duration: 7200

Storage Capabilities

• Time-Series Buffer: Circular buffer for recent high-resolution data
• Store-and-Forward Queue: Persistent queue for unsent messages during connectivity loss
• Local Time-Series Database: Embedded TSDB for extended local storage and querying
• Selective Persistence: Tiered storage based on data importance
• Compacted Storage: Progressively compressed historical data

Synchronization Strategies

• Incremental Sync: Transmitting only new or changed data
• Prioritized Sync: Sending critical data first when connectivity is restored
• Bandwidth-aware Sync: Adjusting synchronization rate based on available bandwidth
• Conflict Resolution: Handling conflicting updates during offline periods

Storage Configuration Example

# Local Storage Configuration

storage:
  # Time-Series Database
  tsdb:
    engine: "sqlite-tsdb"
    path: "/var/lib/edge-node/tsdb"
    max_size: "10GB"
    auto_vacuum: true
    retention:
      default: "7d"
      high_priority: "30d"
      low_priority: "2d"
    downsampling:
      - interval: "1h"
        retention: "90d"
        aggregations: ["avg", "min", "max"]
      - interval: "1d"
        retention: "365d"
        aggregations: ["avg", "min", "max"]
  
  # Store-and-Forward Queue
  message_queue:
    engine: "lmdb"
    path: "/var/lib/edge-node/queue"
    max_size: "2GB"
    max_message_size: "1MB"
    persistence: true
    
  # Synchronization
  sync:
    strategy: "prioritized"
    priorities:
      - name: "critical"
        max_age: "none"  # never expire
      - name: "high"
        max_age: "30d"
      - name: "normal"
        max_age: "7d"
      - name: "low"
        max_age: "2d"
    bandwidth_limits:
      normal: "5Mbps"
      restricted: "1Mbps"
      minimal: "100kbps"

Technical Implementation

• Local Data Storage: Storing sensitive data within the jurisdiction's boundaries
• Data Minimization: Collecting and transmitting only necessary data
• Data Categorization: Classifying data based on sensitivity and applying appropriate controls
• Anonymization at Source: Removing personally identifiable information at the edge
• Federated Analytics: Performing analytics locally and sharing only aggregated results
• Cryptographic Controls: Encryption, tokenization, and secure multi-party computation

Compliance Framework

• GDPR Alignment: Edge processing designed to support data protection principles
• Data Processing Agreements: Clear documentation of data flows and processing activities
• Audit Trails: Comprehensive logging of data access and processing
• Data Subject Rights: Technical capabilities to support access, rectification, and deletion requests
• Privacy by Design: Data protection integrated into edge node architecture
• Data Transfer Mechanisms: Compliant methods for cross-border data transfers when necessary

Network Optimization

• Edge Placement Strategy: Optimizing edge node locations based on sensor density and criticality
• Quality of Service (QoS): Traffic prioritization for latency-sensitive data
• Protocol Selection: Using lightweight protocols optimized for constrained environments
• Connection Pooling: Maintaining persistent connections to reduce handshake overhead
• Compression: Adaptive compression based on data type and network conditions

Processing Optimization

• Workload Placement: Allocating processing tasks to appropriate edge tiers
• Predictive Processing: Anticipating processing needs based on patterns and context
• Resource Allocation: Dynamic allocation of compute resources based on priority
• Parallel Processing: Distributing workloads across available cores and nodes
• Hardware Acceleration: Utilizing specialized hardware for specific tasks (GPU, FPGA)

Pre-Deployment

• Site Survey: Evaluate physical location, power availability, network connectivity, and environmental conditions
• Network Assessment: Measure baseline network performance, identify potential bottlenecks
• Regulatory Compliance: Verify compliance with local regulations (power, RF emissions, permits)
• Security Assessment: Evaluate physical and cyber security requirements
• Capacity Planning: Calculate processing, storage, and bandwidth requirements
• Backup & Recovery: Define backup procedures and disaster recovery plans

Deployment & Validation

• Hardware Installation: Secure mounting, proper grounding, weather protection
• Network Configuration: Secure connectivity, firewall rules, QoS settings
• Software Deployment: OS hardening, application installation, configuration
• Security Implementation: Certificate deployment, access controls, encryption
• Testing & Validation: Functionality testing, performance benchmarking, security testing
• Documentation: As-built documentation, configuration records, test results