2018-自动驾驶移动网络的关键场景（英文版）-5mb

Summary of "Key Scenarios of Autonomous Driving Mobile Network"

Core Content

This document outlines Huawei's vision and approach for achieving autonomous driving in mobile networks, emphasizing the need for scenario-based automation to address the growing complexity of 5G networks and reduce operational costs (OPEX). It identifies seven key scenarios and five core automation capabilities that are essential for the evolution of autonomous mobile networks.

Main Viewpoints

• Automation is critical for the 5G era: With the increasing complexity of mobile networks, automation is necessary to manage operations efficiently and reduce costs.
• Scenario-based automation is the way forward: Huawei advocates for a step-by-step implementation of automation, starting with existing 4G networks and evolving into more advanced 5G capabilities.
• AI is central to automation: AI and machine learning (ML) are key technologies that enable the transformation of mobile networks into autonomous systems.
• Five levels of automation: From manual operations (L0) to full autonomy (L5), Huawei defines a roadmap for network automation that aligns with the development of autonomous vehicles.
• Five core automation capabilities: Programmable, Online, Bridging, Sensibility, and Intelligence are essential to support various automation scenarios and their evolution.

Key Scenarios

The document highlights seven key scenarios in the mobile network lifecycle that are suitable for automation:

1. Base Station Deployment
2. Feature Deployment
3. Network Performance Monitoring
4. Fault Analysis and Handling
5. Network Performance Improvement
6. Home Service Provisioning
7. Power Saving

Each scenario is analyzed in terms of its current status, automation classification, and required capabilities.

5.1 Base Station Deployment

• Definition: The process from site survey to site acceptance.
• Automation Levels:
  • L1: Some elements automated, manual configuration and acceptance.
  • L2: Hardware detection and configuration automated, simplified data.
  • L3: End-to-end (E2E) automation, including self-acceptance.
• Industry Status: Currently at L1-L2, with some platforms approaching L3.
• Key Capabilities:
  • Programmable: Customizable deployment based on policies.
  • Online: All steps completed remotely.
  • Bridging: Unified workflow across systems.
  • Sensibility: Automatic detection of site data.
  • Intelligence: Scenario recognition and wireless parameter planning.

5.2 Feature Deployment

• Definition: Matching the best features to the network based on scenario analysis.
• Automation Levels:
  • L1: Manual feature selection based on experience.
  • L2: Automatic configuration generation and one-button activation.
  • L3: E2E closed-loop feature launch with self-identification and self-acceptance.
• Industry Status: Mostly at L2, with some platforms at L3.
• Key Capabilities:
  • Programmable: Customizable launch process.
  • Online: Feature recommendation and acceptance online.
  • Bridging: Integration with planning and O&M systems.
  • Sensibility: Access to real-time data for scenario recognition.
  • Intelligence: AI-based scenario recognition and gain prediction.

5.3 Network Performance Monitoring

• Definition: Monitoring and forecasting network performance to support business goals.
• Automation Levels:
  • L1: Basic KPI monitoring and manual anomaly detection.
  • L2: 3D visualization of network quality and self-generated planning.
  • L3: E2E closed-loop monitoring and planning with automatic recommendations.
• Industry Status: Mostly at L1-L2, with limited E2E capabilities.
• Key Capabilities:
  • Bridging: Integration with planning and fault diagnosis systems.
  • Sensibility: Relies on multi-dimensional data.
  • Intelligence: Business prediction, 3D positioning, and scenario self-evaluation.

5.4 Fault Analysis and Handling

• Definition: Detecting and resolving network faults quickly to ensure reliability.
• Automation Levels:
  • L1: Manual threshold and correlation rules.
  • L2: Automatic alarm correlation and root cause analysis.
  • L3: Closed-loop analysis and handling with self-healing.
  • L4: Proactive troubleshooting based on trend analysis.
• Industry Status: Mostly at L1-L2, with limited L3 and L4 capabilities.
• Key Capabilities:
  • Bridging: Integration with planning and fault diagnosis systems.
  • Sensibility: Requires access to real-time data.
  • Intelligence: AI-based root cause analysis and predictive fault handling.

Key Automation Capabilities

• Programmable: The ability to define and orchestrate workflows for different automation sub-scenarios.
• Online: Remote execution of tasks, reducing the need for on-site visits.
• Bridging: Seamless data flow between different systems and processes.
• Sensibility: Dependence on real-time and multi-source data for automation.
• Intelligence: Utilization of AI to enhance decision-making and predictive capabilities.

Conclusion

Huawei emphasizes that the autonomous driving mobile network is not a single event but a multi-stage evolution, requiring a scenario-oriented approach. The adoption of AI is essential to overcome the complexity of modern networks and enable more efficient and agile operations. The five levels of automation and five core capabilities provide a clear path for operators to move towards full network autonomy, starting with immediate benefits in current 4G networks and extending to future 5G deployments.