AI-Based Channel State Measurement for Accurate 5G Scheduling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In LTE and 5G systems, the feedback of channel state information is inaccurate due to factors like feedback overhead and quantization, leading to inefficiencies in channel resource scheduling.
Innovation Solution
A communication method utilizing an artificial intelligence algorithm model trained with historical uplink and downlink channel information to determine how network devices obtain downlink channel information, with terminal devices adjusting their measurement parameters accordingly to reduce errors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If traditional channel state measurement methods are used, then feedback overhead is reduced, but measurement precision deteriorates due to quantization errors
Solution Approach 1:
The patent introduces an artificial intelligence algorithm model as an intermediary between the uplink channel information and downlink channel information. The network device inputs uplink channel information into the trained AI model to obtain predicted downlink channel information, which then serves as the basis for generating channel state information. This intermediary approach allows the system to achieve accurate channel state measurement without requiring extensive downlink reference signals, thus resolving the contradiction between feedback overhead and measurement precision.
2Measurement precision
If more configuration parameters are sent for channel state measurement, then measurement precision improves, but device complexity increases
Solution Approach 1:
The patent applies preliminary action by pre-training the artificial intelligence algorithm model offline using historical uplink and downlink channel information as training samples. Once trained, the model is stored in the network device. During actual operation, the pre-trained model can directly process incoming uplink channel information to predict downlink channel characteristics without requiring complex real-time configuration parameter adjustments. This preliminary preparation resolves the contradiction by achieving high measurement precision through the pre-trained model while keeping the operational device complexity low.
3Productivity
If channel reciprocity is assumed, then productivity improves through faster scheduling, but reliability deteriorates due to measurement errors
Solution Approach 1:
The patent implements a feedback mechanism where the network device sends the first information (indicating AI model usage) and second information (indicating reporting type) to the terminal device, and the terminal device feeds back channel state information based on the predicted downlink channel information. The network device then uses this feedback along with the AI model predictions to generate the downlink precoding matrix. This closed-loop feedback system resolves the contradiction by continuously refining the channel state information accuracy while maintaining fast scheduling through the AI model's rapid processing capabilities.
Data Source
AI summary
Provided are a method and device for communication. The method comprises: a network device transmits first information to a terminal device, the first information being used for indicating whether the network device acquires downlink channel information via an artificial intelligence algorithm model, and the artificial intelligence algorithm model being constructed by training with past uplink channel information and past downlink channel information serving as samples. A terminal device measures a channel state on the basis of the first information to acquire channel state information.