Wireless Traffic Flow Prediction Using Heavy-Hitter IP Encoding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing traffic flow prediction methods in wireless networks are inadequate for modern radio networks due to reliance on historical data, limited input parameters, and inefficient QoS differentiation, leading to inaccurate predictions and resource wastage.
Innovation Solution
A computer-implemented method using heavy-hitter IP address and port number encodings, combined with link quality metrics, to train a learning model for predicting traffic flow characteristics, enabling fast and accurate predictions with uncertainty quantification.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional traffic flow prediction methods using historical packet statistics are employed, then predictions can be made for established traffic flows, but the method fails to provide accurate predictions for new traffic flows without sufficient observation time and requires time-consuming feature extraction
Solution Approach 1:
The system performs preliminary encoding of IP addresses into compact representations in advance, creating a lookup table of heavy-hitter IP addresses and their encoded forms. This preliminary action enables immediate prediction for new traffic flows without requiring observation time, as the encoded IP addresses can be directly processed by the machine learning model to predict traffic characteristics
Solution Approach 2:
The invention extracts and utilizes only the most critical identifying feature (IP address) for prediction, encoding it into a compact representation. This extraction approach eliminates the need to collect and process multiple packet statistics over time, enabling immediate accurate predictions for both new and established traffic flows
2Productivity
If carrier aggregation is configured to improve throughput performance, then data transmission capacity increases significantly, but Radio Resource Control signaling overhead and configuration time increase
Solution Approach 1:
The system performs preliminary traffic flow prediction using encoded IP addresses to identify UEs with large buffer data before carrier aggregation configuration is triggered. This preliminary assessment ensures that carrier aggregation is only activated when beneficial, reducing unnecessary configuration overhead and time while maintaining high throughput performance when needed
3Productivity
If RRC signaling is used to reconfigure UE for additional carrier, then bandwidth and throughput are significantly improved, but radio resources are consumed and latency is introduced
Solution Approach 1:
The system uses traffic flow prediction based on encoded IP addresses to provide feedback on which UEs are likely to benefit from additional bandwidth. This feedback mechanism enables the network to make informed decisions about resource allocation, activating carrier aggregation only for UEs with large buffer data, thereby optimizing radio resource consumption while maintaining high bandwidth utilization when needed
Data Source
AI summary
Systems and methods related to traffic flow prediction in a wireless network are disclosed. In one embodiment, a computer-implemented method comprises collecting training data comprising Internet Protocol (IP) addresses extracted from packets for traffic flows in a wireless network and one or more actual traffic type related parameters for each of the traffic flows. The method further comprises training heavy-hitter IP address encodings based on the extracted IP addresses and encoding the extracted IP addresses using the trained heavy-hitter IP address encodings. The method further comprises training a traffic type predictor of a traffic flow predictor based on the encoded IP addresses and the one or more actual traffic type related parameters for each of the traffic flows, where the traffic type predictor is a learning model that maps encoded IP addresses to one or more predicted traffic type related parameters.