REAL TIME COMMUNICATIONS NETWORK

Introduction

The AI-PRISM real-time communication network facilitates communication between sensors, the AI-PRISM Data Platform, edge devices, and software components. Its goal is to ensure high-quality service while remaining adaptable. It employs advanced industrial IoT technologies for flexible wireless sensing combined with industry-integrated software-defined networks (SDN). SDN centralizes network management by offering a comprehensive view across various technologies, such as Time Sensitive Networks, Ethernet, Wi-Fi, and IEEE 802.15.4. This centralized approach, overseen by the SDN controller, optimizes end-to-end service quality. The SDN controller manages multiple network technologies, simplifying device configurations. This relaxes application development in the fog as the controller provides high-level APIs, connecting apps to networks without complex configurations.

As highlighted in the architecture figure, this infrastructure is the enabler to achieving real-time communications in AI-PRISM. It has a stack of protocols that allow centralized control of all node processes (e.g. Medium Access Control - MAC, Routing, QoS, Application). As mentioned above, this is based on the SDN paradigm, which manages networks by separating the control and data planes. The control plane is entirely software-controlled, where a centralized agent, in this case, an SDN controller, decides to configure resources and rules in the nodes. The data or forwarding plane are made up of the nodes, which operate in plug-and-play mode, as the controller sends each node all the necessary instructions to achieve maximum performance. In addition, the SDN controller is used as an integration point for the physical network infrastructure and the application that collects the network information and requests network reconfiguration. For that, use a specific protocol for each network technology, for example, OpenFlow for wired SDN network, NetFlow for TSN, and even non-standard protocols such as SDN WISE for WSN.

SDN architecture for Real-time communication network

Key benefits of this setup include:

Technical Specifications

The AI-PRISM project aims to establish a network architecture that supports robust, real-time, and secure interactions among its components, emphasizing quality of service, adaptability, mobility, and scalability.

  1. SDN Paradigm: The SDN model governs networks by distinctly managing the control and data planes.

  2. Control Plane: Governed by software, this is where the SDN controller resides, which dictates resource configurations and node rules.

  3. Data/Forwarding Plane: Comprising nodes, they operate plug-and-play, receiving directives from the controller for optimal performance.

  4. Integration with SDN Controller: As a nexus for physical network infrastructure and applications, it acquires network data and prompts network recalibrations. It uses specific protocols tailored to each network tech, such as OpenFlow for wired networks and SDN WISE for WSNs.

  5. Software Defined Wireless Sensor Network (SD-WSN): This deterministic system upholds industrial benchmarks using wireless communication tools devised by ITI, which spearheads the SDN/NFV Network Infrastructure development. Central to AI-PRISM's real-time communication capabilities, the infrastructure (illustrated in Figure 4) incorporates a set of protocols facilitating centralized management of node processes, such as Medium Access Control, Routing, and QoS.

  6. Base Layer: This prioritizes developing and integrating communication technologies for live interactions. Whether wired or wireless, it's foundational in marrying Information Technology (IT) and Operational Technology (OT) flows. The architecture prioritizes real-time, resilient, and safe communications in the AI-PRISM's decentralized system, tackling the complexity of adaptability, mobility, and scalability in envisioned scenarios. Key drivers here are industrial-grade IoT technologies, a flexible wireless sensing layer, and the integration of software-defined networks in the industry.

  7. Upper Layer: The SDN Controller furnishes protocols such as MQTT and KAFKA alongside a management API. This permits the AI-PRISM Data Platform to outline and evolve all its functions. The platform offers a comprehensive data model and interfaces granting access to data assets, like data broker topics and repository logs, as well as functions for data governance.

Technical specifications in SDN architecture for Real-time communication network

Software and hardware Requirements

Hardware Requirements
Software Requirements

Usage Manual

In this section the main features of the component should be identified and explained.

It also should include the explanation step by step each of the identified use cases.

Use Case 1.

Use Case Diagram

Through this usage model, the AI-PRISM Real-Time Communications Network (RC) component facilitates seamless interaction between diverse agents and the SDN controller, enabling dynamic network management and ensuring optimal QoS for all users and applications.

User requests: End-users or applications initiate communication requests, specifying their QoS requirements, such as bandwidth, latency, or priority levels.

Agent communication: Network agents, such as IoT devices, sensors, and routers, relay the user requests and report their current network conditions, including link capacity, utilization, and congestion, to the SDN controller.

SDN controller processing: The SDN controller analyses the user requests and network conditions, considering the QoS requirements and any predefined policies or rules. Based on this analysis, the controller determines the optimal flow configurations and routing paths to meet the QoS demands.

Flow configuration and routing update: The SDN controller communicates the new flow configurations and routing paths to the relevant network agents, instructing them to update their forwarding tables and apply the required changes.

Network agents' execution: update their forwarding tables and implement the new flow configurations and routing paths as directed by the SDN controller.

Monitoring and adjustment: The SDN controller continuously monitor network performance and user demands, adjusting flow configurations and routing paths in real-time to maintain the desired QoS levels and adapt to network changes or fluctuations.

Use case for SDN communication network

Use Case Mock-ups

The dashboard offers a centralized interface for efficiently managing and monitoring software-defined networks. It encompasses several key features.

-Flow configuration: Allows administrators to define and set up data flows, routing policies, traffic prioritization, and security rules.

-Device statistics: Provides detailed insights on network devices' performance, resource usage, and errors, enabling proactive problem identification and resolution.

-Network topology visualization: Displays a graphical representation of the entire network topology, highlighting interconnected devices and links, identifying bottlenecks and failure points easier.

-Resource monitoring: Offers real-time information on network resource usage and availability, ensuring efficient allocation and prioritization of critical services.

-Data flow path analysis: Enables administrators to track and analyse data flow paths in the network, identifying performance or congestion issues and implementing more efficient routing solutions.

SDN controller Mock-ups

Functional Specifications

Functional Block Diagram

The SDN network focused on QoS and Dynamic reconfiguration is a real-time network infrastructure that uses a monitoring system to collect data from the network devices. The monitoring system uses a southbound interface to communicate with the network devices and to request specific data that is relevant to the QoS and Dynamic reconfiguration requirements. The monitoring system can also modify the network behavior by sending commands to the network devices through the southbound interface.

The monitoring system monitors the global view of the network, resources, and requirements. It also monitors the real-time network conditions and traffic trends. The monitoring system sends the monitoring data to other systems through northbound interfaces. The monitoring system also uses the data from the industrial process to understand the behavior of the process and to identify potential problems.

The advantage of this network is that it can integrate multiple components through the southbound interface and abstract their behavior using APIs. This makes it easy to manage and monitor the network, even when it is complex and heterogeneous. The SDN network is also focused on QoS and Dynamic reconfiguration. This means that it can be used to ensure that the network provides the required level of service for different applications and traffic flows. It can also be used to reconfigure the network dynamically in response to changes in traffic conditions or to meet new requirements.

Functional specifications in SDN architecture for Real-time communication network

Main interfaces

List of main interfaces between functional components shown in the figure.

ID Component Name Description Sense
1 Real-time network infrastructure Network Devices Physical network being monitored including routers, switches, firewalls, managed by a SDN network controller. In
2 Start real-time communication Data Collection Monitoring system begins to collect data from network devices using a southbound interface. In
3 Real-time data from devices Data Collection Data collected from network devices, including traffic statistics, device status, event logs. In
4 Data flow requirements (API) Data Request API allowing monitoring system to request specific data from network devices relevant to QoS and Dynamic reconfiguration. In
5 Monitoring Network Health Monitoring Monitoring the overall health of the network, resources, and requirements of network devices to identify and correct problems. In
6 Flow Monitoring Traffic Monitoring Monitoring real-time conditions and traffic trends of the network, such as traffic levels, deadlines, delays, and packet loss. In
7 Send data through Northbound interfaces Data Sending Sending monitoring data to other systems, such as a SCADA system or a cloud-based monitoring system. Out
8 Industrial Process Input Process Monitoring Input from the industrial process being monitored, such as sensor data and control commands. In

Sequence Maps

The process shows how SDN can monitor and control a network in an industrial process. The process begins with the collection of data from the network devices. This data is then used to create a real-time view of the network. The monitoring system can then use this data to identify potential problems and take corrective action. The monitoring system can also use the data to dynamically reconfigure the network in response to changes in traffic conditions or to meet new requirements. For example, if the monitoring system detects a lot of traffic on a particular link, it can dynamically reroute the traffic to a different link. This can help ensure the network is always performing at its best. This SDN network can integrate with other systems, such as an ERP or cloud-based monitoring system. This integration can help to provide a more comprehensive view of the industrial process and improve the efficiency of the monitoring and control process.

interactions between the functional componentes in SDN architecture