Architecture Overview

The Zenoh Multi-Protocol Bridge is designed with a microservices architecture, where each component is containerized and communicates over well-defined protocols.

System Architecture

The system follows a hub-and-spoke architecture with the Zenoh Router at its core, enabling seamless communication between different protocol domains.

graph TB
    subgraph "Docker Environment"
        subgraph "MQTT Domain"
            MQTTPub[MQTT Publisher<br/>Auto Data Generator]
            MQTTSub[MQTT Subscriber]
            Mosquitto[Mosquitto Broker<br/>:1883, :9001]
        end

        subgraph "Zenoh Core"
            ZenohRouter[Zenoh Router<br/>:7447, :8000<br/>Central Hub]
        end

        subgraph "Bridge Layer"
            MQTTBridge[Zenoh-MQTT Bridge<br/>:8001, :1884<br/>Protocol Translation]
            ROSBridge[Zenoh-ROS2DDS Bridge<br/>:8002, :7449<br/>Protocol Translation]
        end

        subgraph "ROS2 Domain"
            ROS2[ROS2 Humble<br/>:8765<br/>Foxglove Bridge]
            ROS2Pub[ROS2 Publisher<br/>/chatter topic]
        end

        subgraph "Zenoh Clients"
            ZenohSub[Zenoh Subscriber<br/>Python Application]
        end

        subgraph "Management"
            NodeRED[Node-RED<br/>:1880<br/>Monitoring & Control]
        end
    end

    subgraph "External"
        ExternalPub[pub.py<br/>Host Zenoh Publisher]
        ExternalSub[sub.py<br/>Host Zenoh Subscriber]
        WebBrowser[Web Browser<br/>User Interface]
    end

    %% Connections
    MQTTPub -->|MQTT Publish| MQTTBridge
    MQTTBridge <-->|MQTT| Mosquitto
    MQTTBridge <-->|Zenoh Protocol| ZenohRouter

    Mosquitto <-->|MQTT Subscribe| MQTTSub
    Mosquitto <-->|MQTT| NodeRED

    ZenohRouter <-->|Zenoh Subscribe| ZenohSub
    ZenohRouter <-->|Zenoh Protocol| ROSBridge

    ROSBridge <-->|DDS Communication| ROS2
    ROS2Pub -->|ROS2 Topic| ROS2

    NodeRED -.->|Monitor| MQTTBridge
    NodeRED -.->|Monitor| ROS2

    ExternalPub <-->|TCP:7447| ZenohRouter
    ExternalSub <-->|TCP:7447| ZenohRouter

    WebBrowser -->|HTTP| NodeRED
    WebBrowser -->|WebSocket| ROS2

    classDef mqtt fill:#e1f5fe
    classDef zenoh fill:#f3e5f5
    classDef bridge fill:#fff3e0
    classDef ros fill:#e8f5e8
    classDef mgmt fill:#fce4ec

    class Mosquitto,MQTTPub,MQTTSub mqtt
    class ZenohRouter,ZenohSub zenoh
    class MQTTBridge,ROSBridge bridge
    class ROS2,ROS2Pub ros
    class NodeRED mgmt

Design Principles

1. Separation of Concerns

Each component has a single, well-defined responsibility:

• Protocol Bridges: Handle translation between protocols
• Message Brokers: Manage message routing within their protocol domain
• Core Router: Provides the central Zenoh routing infrastructure
• Clients: Consume or produce data in specific protocols

2. Scalability

The architecture supports horizontal scaling:

• Multiple bridge instances can connect to the same Zenoh router
• Clients can be added or removed dynamically
• Load can be distributed across multiple routers

3. Loose Coupling

Components are loosely coupled through:

• Standardized protocols (MQTT, Zenoh, DDS)
• Configuration-driven topic mapping
• Docker networking for service discovery

4. Protocol Agnostic

The system abstracts protocol-specific details:

• Applications don't need to know about all protocols
• New protocols can be added with new bridge services
• Protocol translation is transparent to clients

Key Components

Core Infrastructure

Zenoh Router

• Role: Central message routing hub
• Technology: Eclipse Zenoh
• Key Features:
• High-performance pub/sub
• Store/Query capabilities
• REST API for monitoring
• Zero-copy architecture

Docker Network

• Name: zenoh-mqtt-net
• Type: Bridge network
• Purpose: Container-to-container communication
• DNS: Automatic service name resolution

Protocol Domains

MQTT Domain

• Broker: Eclipse Mosquitto
• Clients: Publishers and Subscribers
• Use Cases: IoT devices, sensors, legacy systems

ROS2 Domain

• Framework: ROS2 Humble
• Middleware: CycloneDDS
• Use Cases: Robotics, autonomous systems

Zenoh Domain

• Router: Eclipse Zenoh Router
• Clients: Python subscribers, external scripts
• Use Cases: Edge computing, distributed systems

Bridge Services

Zenoh-MQTT Bridge

• Function: Bidirectional MQTT ↔ Zenoh translation
• Configuration: JSON5 based
• Features:
• Topic filtering
• Scope-based key mapping
• Generalised subscriptions

Zenoh-ROS2DDS Bridge

• Function: Bidirectional ROS2 ↔ Zenoh translation
• Configuration: JSON5 based
• Features:
• Namespace isolation
• Topic/service/action bridging
• Selective filtering

Management & Monitoring

Node-RED

• Purpose: Visual programming and monitoring
• Capabilities:
• Flow-based programming
• Real-time dashboards
• Protocol integration
• Rule-based automation

Communication Patterns

Pub/Sub Pattern

The system primarily uses publish-subscribe pattern:

sequenceDiagram
    participant Pub as Publisher
    participant Router as Zenoh Router
    participant Sub1 as Subscriber 1
    participant Sub2 as Subscriber 2

    Sub1->>Router: Subscribe to "sensor/*"
    Sub2->>Router: Subscribe to "sensor/temperature"

    Pub->>Router: Publish "sensor/temperature" = 25.5

    Router->>Sub1: Deliver (matches "sensor/*")
    Router->>Sub2: Deliver (matches "sensor/temperature")

Request/Response Pattern

ROS2 services use request/response:

sequenceDiagram
    participant Client as ROS2 Client
    participant Bridge as ROS2 Bridge
    participant Router as Zenoh Router
    participant Service as ROS2 Service

    Client->>Bridge: Service Request
    Bridge->>Router: Zenoh Query
    Router->>Bridge: Route to Service
    Bridge->>Service: DDS Service Call
    Service->>Bridge: Service Response
    Bridge->>Router: Zenoh Reply
    Router->>Bridge: Route to Client
    Bridge->>Client: Service Response

Data Flow Layers

Layer 1: Data Generation

• MQTT publishers generate sensor data
• ROS2 nodes publish robot data
• External scripts publish test data

Layer 2: Protocol Translation

• MQTT Bridge translates MQTT ↔ Zenoh
• ROS2 Bridge translates DDS ↔ Zenoh
• Bidirectional data flow

Layer 3: Routing

• Zenoh Router distributes messages
• Pattern-based subscription matching
• Efficient message delivery

Layer 4: Data Consumption

• MQTT subscribers receive IoT data
• ROS2 nodes receive robot commands
• Zenoh subscribers get unified data stream
• Node-RED visualizes and processes data

Configuration Management

Static Configuration

• Docker Compose orchestration
• Service configurations (JSON5 files)
• Network and volume definitions

Dynamic Configuration

• Runtime topic subscriptions
• Flow modifications in Node-RED
• Adjustable publish rates

Security Considerations

Network Isolation

• Services communicate within Docker network
• Port mapping controls external access
• Host access restricted to necessary ports

Access Control

• Mosquitto ACL configuration
• Topic filtering in bridges
• Namespace isolation in ROS2

Future Enhancements

• TLS/SSL for encrypted communication
• Authentication tokens
• Fine-grained access policies

Deployment Architecture

Single-Host Deployment (Current)

┌─────────────────────────────────────┐
│         Docker Host                 │
│  ┌───────────────────────────────┐  │
│  │   zenoh-mqtt-net (bridge)     │  │
│  │  ┌──────┐  ┌──────┐  ┌──────┐│  │
│  │  │MQTT  │  │Zenoh │  │ROS2  ││  │
│  │  │Domain│  │Core  │  │Domain││  │
│  │  └──────┘  └──────┘  └──────┘│  │
│  └───────────────────────────────┘  │
└─────────────────────────────────────┘

Multi-Host Deployment (Future)

┌──────────────┐      ┌──────────────┐      ┌──────────────┐
│   Host 1     │      │   Host 2     │      │   Host 3     │
│ ┌──────────┐ │      │ ┌──────────┐ │      │ ┌──────────┐ │
│ │MQTT      │ │◄────►│ │Zenoh     │ │◄────►│ │ROS2      │ │
│ │Domain    │ │      │ │Router    │ │      │ │Domain    │ │
│ └──────────┘ │      │ └──────────┘ │      │ └──────────┘ │
└──────────────┘      └──────────────┘      └──────────────┘

Performance Characteristics

Latency

• MQTT → Zenoh: < 10ms (typical)
• ROS2 → Zenoh: < 5ms (typical)
• End-to-End: < 20ms (MQTT → Zenoh → ROS2)

Throughput

• MQTT Bridge: 10K+ messages/second
• Zenoh Router: 1M+ messages/second
• ROS2 Bridge: 50K+ messages/second

Resource Usage

• Memory: ~500MB total (all services)
• CPU: < 5% idle, < 30% under load
• Network: Minimal latency, high efficiency

Next Steps

• System Components - Detailed component descriptions
• Data Flow - Message flow patterns
• Network Architecture - Network topology and configuration