ADN

What Is ADN?

ADN (Autonomous Driving Network) is a technology that uses the power of AI to independently understand service intentions and network goals. It surpasses the limitations of manual processing capabilities and introduces transformative advancements. By leveraging data computing, ADN compensates for blind spots in operations and maintenance, which go beyond the scope of manual experience. Through continuous machine learning, ADN transcends the constraints of decision-making based solely on human experience. As a result, ADN achieves automation, self-healing, self-optimization, and autonomy, paving the way for the development of intelligent networks that are pervasive and ubiquitous.

The Significance of ADN

With the rapid advancement of an intelligent society, where everything is interconnected and intelligent, the demands for network capabilities have evolved significantly. People now expect more than just multi-device collaboration; they seek ubiquitous access for all individuals, devices, applications, and terminals. Consequently, this has led to increased expectations and higher requirements for network performance.

In the context of enterprises, the acceleration of digitalization necessitates the rapid iteration of services. This involves configuring networks across campuses, WANs, data centers, and multiple clouds, surpassing the boundaries of network operations and maintenance (O&M). Manual O&M struggles to keep up with the frequent changes required to adapt to the growing number of critical services. To address this, enterprises require intelligent detection and self-optimization capabilities, enabling a shift from static policies to dynamic policies.

On an individual level, networks have become deeply ingrained in various aspects of our work and personal lives. Activities like live streaming, shopping festival participation, and online courses are now part of our daily routines. The unpredictable bursts in network traffic and the need for flexible allocation of network resources have become essential. Additionally, the rising popularity of IoT, smart homes, and the Internet of Vehicles (IoV) introduces new challenges in terms of network stability and security.

To overcome these challenges and deliver an exceptional network experience with enhanced service agility and reduced operational expenditure (OPEX), enterprises require an ADN solution that integrates AI and big data technologies. Such a solution enables network automation and intelligent O&M, empowering enterprises to meet evolving demands effectively.

The triple-layer AI architecture of ADN

1. Cloud + Al: Leveraging a cloud platform, the system offers data lakes, model training, ecosystem openness, and developer services. Through continuous learning and evolution, AI enhances the system's intelligence, enabling it to adapt to rapid service and network changes. The goal is to deliver improved services and a superior user experience.

2. Network + Al: Intents at the service layer drive the automatic generation and deployment of network configurations. This ensures that the network consistently aligns with service intents. Integrating management, control, and analysis, this layer enables real-time closed-loop network management and control.

3. NE + Al: Network devices support edge inference and real-time decision-making using AI Turbo. They dynamically adjust forwarding policies based on service intents, ensuring an optimal real-time service experience.

How Are ADN Levels Categorized?

ADN establishes clear criteria for level setting and outlines an evolution roadmap to assist users in their digital transformation journey. These guidelines inform users about the necessary actions and timing required to achieve successful transformations. Moreover, ADN offers a range of solutions and an optimal technical architecture tailored to each phase. This guidance empowers industries to attain the ultimate network experience and supports enterprises in executing their digital transformation and upgrade strategies. By embracing ADN, organizations can maximize efficiency, minimize costs, and foster continuous innovation throughout their digital journey.

There are six levels of network governance:

• L0: This level involves standard operation processes where network operations and maintenance (O&M) rely entirely on manual tasks.

• L1: At this level, a certain degree of automation is introduced, allowing network O&M to be carried out with the assistance of tools.

• L2: Online data processing is implemented, enabling specific functions to be automatically executed without manual intervention in certain modules.

• L3: Operations driven by data are performed, granting the network a level of autonomy under specific conditions.

• L4: Intelligent processing of data is achieved, leading to a high degree of network autonomy. Automatic network O&M can be accomplished in most common scenarios.

• L5: This level represents the pinnacle of network governance, where data self-driving is present in all scenarios. The network achieves full autonomy, with the capability for automatic inference, learning of new scenarios, and freeing people from manual tasks.