Predictive Buffer State Reporting for Low-Latency Uplink Grants
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing communication systems face latency issues in data transmission due to the need for terminal devices to wait for data assignment to logic channels before calculating buffer sizes for buffer state reports (BSR), which prolongs the time required for data transmission to network nodes.
Innovation Solution
Implementing a mechanism where terminal devices predict buffer sizes associated with data to be transmitted and transmit a buffer state report (BSR) including the predicted size to network nodes, using artificial intelligence algorithms like LSTM or Arima machine learning models, thereby reducing the need for data assignment to logic channels.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If terminal devices wait for data assignment to logic channels before calculating buffer sizes, then measurement precision of buffer state is improved, but loss of time increases
Solution Approach 1:
The terminal device performs preliminary prediction of the buffer size using machine learning models (LSTM, Arima, or reinforcement learning) before the data is actually assigned to logic channels. This allows the BSR to be transmitted earlier in the process, reducing the time latency while the prediction algorithms estimate the buffer state based on available information such as data size, type, content, required transmission rate, application behavior, user behavior, radio conditions, and traffic patterns.
2Productivity
If terminal devices use prediction algorithms for buffer size, then productivity of data transmission is improved, but device complexity increases
Solution Approach 1:
The patent introduces machine learning models (LSTM, Arima, reinforcement learning) as intermediary components that act as predictors between the raw data characteristics and the buffer size calculation. These models process input features such as data size, type, content, required transmission rate, application behavior, user behavior, radio conditions, and traffic patterns to generate predicted buffer size estimates, thereby simplifying the overall process while improving productivity.
3Loss of time
If terminal devices transmit BSR with predicted buffer size, then loss of time is reduced, but reliability of buffer state information may worsen
Solution Approach 1:
The system implements feedback mechanisms where the terminal device continuously monitors actual buffer sizes and compares them with predicted values. This feedback is used to refine and update the machine learning models (LSTM, Arima, reinforcement learning) over time, improving the accuracy of predictions. The terminal device can also adjust prediction parameters based on observed discrepancies between predicted and actual buffer states, thereby maintaining reliability while achieving early transmission.
Data Source
AI summary
Embodiments of the present disclosure provide methods and apparatus for buffer state report. A method performed at a terminal device may comprise: predicting a buffer size associated with data to be transmitted; and transmitting a buffer state report, BSR, including the predicted buffer size to a network node. A method performed at a network node may comprise: receiving a buffer state report, BSR, including a predicted buffer size associated with data to be received from a terminal device; and transmitting a grant for the data according to the received BSR. The latency of data transmission from the terminal device to the network node may be reduced.