Intelligent Computing

Corporation Consumer Carrier

中文

5G-Advanced

Universal Intelligence

Solution Overview

Scenarios and Problems

Inadequate data

The data in radio access network are too massive yet poorly utilized with too much redundancy, imbalance, and lack of informative interpretation, which makes it difficult to meet the requirements of increasingly demanding applications.

Limited applicability of AI algorithms/models

The conventional way of AI algorithms and models often comes with very limited applicability, and there is hardly any unified framework based on which fundamental capabilities can be leveraged efficiently.

Inadequate computing power

The computing power of wireless base stations, including computing, storage and transmission capabilities, can’t support large-scale AI model training and inter-site sharing of computing power.

Advantages and Highlights

AI introduction in new ways

Introducing AI into radio network in new ways of leveraging data, algorithms and computing power, so as to improve network resource utilization efficiency

Atomic capabilities introduction

Atomic capabilities including data processing and algorithms can be accessed by more intelligent applications, not directly but through a middle AI platform with much better flexibility and efficiency.

Customer Values

Data orchestration on demand

Building distributed data services on top of a cross-domain E2E data model, enabling data orchestration on demand to support AI-based applications

Discrete algorithm services

The flexibly coupled and highly reusable atomic algorithms can be used to offer discrete and algorithm services to RAN applications.

Inter-site computing power sharing

Inter-site computing power sharing helps meet high demands on base station’s AI capability without increasing cost.