Neural Network MAC Scheduling for Variable Terminal Counts
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing MAC layer scheduling algorithms struggle to achieve optimal performance in complex communication systems due to their reliance on closed-form models and formulas, which cannot accurately capture the system's dynamics.
Innovation Solution
A neural network-based scheduling method that uses a neural network model to determine scheduling weights for terminal devices, allowing for flexible adaptation to different numbers of scheduled terminal devices without the need for retraining.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a closed-form model and formula are used for scheduling, then the scheduling algorithm is simple to implement, but it cannot accurately capture the complex system dynamics and achieve optimal performance
Solution Approach 1:
The patent replaces traditional closed-form mathematical models and formulas with a neural network-based scheduling model. The neural network learns complex system dynamics from data and provides accurate scheduling decisions without requiring explicit mathematical formulations, thus resolving the contradiction between accuracy and model complexity.
Solution Approach 2:
The patent transforms the scheduling approach from using fixed closed-form parameters to using learned parameters from neural network training. The neural network adapts its parameters based on system conditions, enabling accurate modeling of complex dynamics while maintaining computational efficiency through the learned parameter representations.
2Reliability
If a neural network model is trained for a specific number of terminal devices, then it achieves optimal performance for that scenario, but it cannot be reused when the number of terminal devices changes
Solution Approach 1:
The patent designs the neural network model with a universal architecture that can handle variable numbers of terminal devices. The model processes input data in a way that adapts to different scenario sizes without requiring retraining, enabling the same model to achieve optimal performance across multiple scenarios with different numbers of terminal devices.
Solution Approach 2:
The patent implements a dynamic model architecture where the neural network can adapt its processing to different input sizes. The model structure allows it to dynamically adjust to varying numbers of terminal devices while maintaining its learned scheduling strategies, thus achieving both optimal performance and reusability.
3Reliability
If multiple neural network models are trained for different numbers of terminal devices, then each model achieves optimal performance for its specific scenario, but the overall system complexity and training time increase
Solution Approach 1:
The patent trains a single universal neural network model that can handle multiple scenarios with different numbers of terminal devices. This eliminates the need for training separate models for each scenario, significantly reducing training time while maintaining optimal performance across all scenarios through the model's ability to adapt to different input configurations.
Data Source
Figure 1
Figure 2~3
Figure 4~5
AI summary
This application provides a scheduling method and apparatus, to optimize a MAC scheduling algorithm. The method includes: a network device uses a scheduling model applicable to K users for one or more times, and determines a scheduled terminal device based on scheduling weights output by the scheduling model for one or more times. Each time the scheduling model is used, status information of K terminal devices in n to-be-scheduled terminal devices is input to the scheduling model, and the scheduling model outputs scheduling weights respectively corresponding to the K terminal devices, where K is an integer greater than 1, and n is an integer greater than K. Therefore, regardless of a quantity of to-be-scheduled terminal devices in a communication system, the network device may reuse the scheduling model without re-establishing a scheduling model, so that the scheduling model is expandable and applicable to scenarios in which quantities of to-be-scheduled terminal devices are different.