Neural Network MCS Selection Under Delayed Channel Feedback
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing MCS selection methods in wireless communication are hindered by feedback delay, measurement errors, and quantization errors, leading to suboptimal channel transmission performance due to deviations from actual channel quality.
Innovation Solution
A neural network-based block error rate prediction model is used to predict the block error rate for each MCS, allowing for real-time selection of an optimal MCS based on channel parameters, thereby improving transmission performance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If traditional AMC technology is used to select MCS based on channel quality feedback, then transmission rate can be increased to maximize channel capacity, but measurement errors and feedback delay cause the selected MCS to deviate from actual channel quality
Solution Approach 1:
The patent replaces traditional mechanical measurement and feedback systems with a neural network-based prediction system. The neural network model processes channel parameters directly to predict block error rate, eliminating the need for traditional feedback mechanisms and reducing measurement errors. This substitution of the measurement system with an intelligent prediction system resolves the contradiction between achieving high transmission rates and maintaining accurate channel quality measurement.
2Stability of the object's composition
If OLLA adjustment algorithm is used to correct channel quality indicator, then channel quality indicator stability is improved, but preset target block error rate deviates from optimal value leading to low transmission performance
Solution Approach 1:
The patent changes the parameter used for MCS selection from a stability-focused channel quality indicator (corrected by OLLA) to a prediction-focused block error rate predicted by the neural network. This parameter change allows the system to achieve both stability and optimal transmission performance, as the neural network's prediction directly targets the optimal block error rate without the deviations introduced by preset target values in traditional OLLA algorithms.
3Productivity
If neural network model is used to predict block error rate for each MCS, then transmission performance is improved through accurate MCS selection, but model complexity and computational requirements increase
Solution Approach 1:
The patent applies preliminary action by pre-training the neural network model using historical channel data and block error rate information. This pre-training phase captures the complex relationships between channel parameters and block error rates, allowing the model to make accurate predictions during actual transmission without requiring complex real-time computations. The preliminary learning phase resolves the contradiction by preparing the model in advance, reducing operational complexity while maintaining high transmission performance.
Data Source
AI summary
This application provides a modulation and coding scheme (MCS) selection method. In the method, a first communication device determines a predicted block error rate that corresponds to each MCS of a plurality of MCSs at a transmission time interval (TTI) 1 by using a neural network model, where a prediction parameter that corresponds to each MCS includes a channel parameter and the MCS. The first communication device selects, from the plurality of MCSs based on the plurality of MCSs and the predicted block error rate that corresponds to each MCS, a target MCS that corresponds to the TTI 1, and sends data to a second communication device at the TTI 1 based on the target MCS. The method disclosed herein helps improve channel transmission performance.