Huawei releases SoftCOM white paper on autonomous driving telecommunications network intelligence

[Shenzhen, China, April 17, 2019] Today, during the Huawei 2019 Analyst Conference held in Shenzhen, Huawei officially released the SoftCOM Communication Intelligence White Paper for Autonomous Driving Telecom Networks. It explains the strategic evolution, architecture innovation, and hierarchical concept of autonomous driving telecom networks, demonstrates the platform value, scenario exploration, and case implementation of telecom network AI, and points out the direction for the application of AI in autonomous driving telecom networks.

With the rapid increase in network complexity, the trend of network automation has become an industry consensus. In combination with the development history of telecommunications networks, Huawei's network architecture proposed the Single strategy in the ALL IP stage. After the rise of cloud computing, it entered the ALL Cloud stage in 2012 and proposed SoftCOM (Software defined Network + TeleCOM) to realize a data center-centric network. In recent years, with the development of artificial intelligence technology, ALL Intelligence has been proposed, artificial intelligence has been introduced into telecommunications networks, SoftCOM AI was born, and a self-driving telecommunications network was built. After Huawei released the concept of self-driving networks, it borrowed the hierarchical concept of automobile self-driving (L0-L5), starting from the core operating processes of operators, through the explanation of different scenarios and hierarchical definitions, it aims to gather industry consensus, so that every step of effort can be effectively combined with the dream of realizing fully autonomous driving, and effectively promote the realization of network automation.

White Paper*** Released Huawei SoftCOM Communication Intelligent Platform Series: Data Lake, Training Platform and Inference Framework, as well as the telecom network AI services that can be provided based on Huawei SoftCOM Communication Intelligent Platform Series. SoftCOM Communication Intelligent Data Lake provides thematic and training data set services. Accurately adapt to various model training needs, combine domain expert experience and knowledge and live network feedback to provide data governance services. SoftCOM Communication Intelligent Training Platform integrates key features such as telecom field feature processing and auxiliary rapid identification, and has built-in typical AI algorithms in the telecom field, such as anomaly detection, root cause analysis, business prediction, etc., to achieve one-stop efficient model training services. SoftCOM Communication Intelligent Inference Framework supports automatic installation, update and deployment of models, grayscale release, cross-network data collection and transmission, multi-model inference execution, timed scheduling and effective inference evaluation, etc., to achieve agile inference services.

The white paper proposes the scenarios and rhythm of introducing AI from three levels: infrastructure layer, network equipment layer, and service bearer layer. The application of AI in the infrastructure layer mainly provides AI accelerators for active hardware facilities to achieve training and reasoning capabilities at different levels. The application of AI in the network facility layer realizes intelligent network optimization, operation and maintenance, and management and control of networks and services. It realizes network KPI optimization, routing optimization, network policy optimization, etc. The application of AI in the service bearer layer, with the deployment of virtualized networks, can gradually add AI capabilities to the orchestration layer, improve the automation and intelligence level of service orchestration and end-to-end resource orchestration, and achieve service quality optimization.

At the same time, the white paper*** announced Huawei's exploration and planning of combining 5G and AI. Huawei will integrate AI into the end-to-end network and realize intelligent 5G by combining the planning, construction, optimization and operation and maintenance of 5G networks throughout the entire life cycle. Practice is carried out from three aspects: 5G site construction intelligence, 5G slicing intelligence and 5G DC intelligence. 5G site construction automatically recommends site locations and site opening parameters based on multi-target site planning, automatically activates large-granularity features, and automatically optimizes parameters. 5G slicing intelligence continuously enhances network slicing based on the automation capabilities built on the central cloud/edge cloud, realizes real-time resource perception of services, adjusts on demand, and shortens service launch time with minimal configuration. 5G DC intelligence includes intelligent expansion configuration, cross-layer linkage energy saving, intelligent inspection, hard disk failure prediction and maintenance, MEC intelligent operation and maintenance, etc.

Han Yufa, President of Huawei SoftCOM Solutions, said: "Autonomous driving telecommunications networks are a new mission for the industry, which requires systematic thinking and innovation from the architecture and key technology levels. SoftCOM Communication Intelligent Solutions introduces AI in all aspects of the end-to-end process of network planning, construction, maintenance, and optimization through architectural innovation, enabling full-scenario autonomous driving telecommunications networks."

[[262822]]

Han Yufa, President of SoftCOM Solutions, officially released the SoftCOM Communication Intelligence White Paper on Autonomous Driving Networks

"The intelligentization of operator networks cannot be achieved overnight; it is a long-term process. SoftCOM AI is the implementation of Huawei's All Intelligence strategy in the telecommunications field. The core AI capabilities are accumulated and grown based on Huawei's long-term and resolute strategic investment in All Intelligence. Combined with scenarios in the telecommunications field, it helps operators build an autonomous driving network that never fails and realize digital and intelligent transformation as soon as possible." Han Yufa said.

White paper download link: Download PDF

Currently, SoftCOM's intelligent communication solution is being tested and is open to industry insiders. Welcome to visit our official cloud service website to experience the telecom AI development platform and intelligent communication services. You can also follow the "Artificial Intelligence Garden" WeChat portal to experience AI model training in the communication field online on your mobile phone.