Handover Prediction Scheduling for Reliable 5G Traffic Delivery
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing 5G wireless communication systems face challenges in managing handover events for mobile devices with varying QoS demands, particularly in ultra-reliable low latency communications (URLLC), due to unpredictable link failures and varying load demands, which can impact network efficiency and power consumption.
Innovation Solution
A method involving a radio network node that receives user equipment (UE) handover predictions, adjusts traffic scheduling based on predicted handover times, and configures UE handover prediction reporting criteria to optimize delivery and reduce link failures.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Use of energy by moving object
If a RAN node activates network energy saving mode to reduce power consumption, then energy efficiency is improved, but traffic delivery reliability deteriorates due to potential link failures during handover
Solution Approach 1:
The RAN node receives handover prediction reports from UEs and proactively identifies traffic that can be delivered before handover occurs. By performing preliminary traffic delivery actions based on predicted handover timing, the system ensures critical traffic is transmitted while the current link is still reliable, thereby maintaining delivery reliability even when entering energy saving mode that might cause link failures during handover.
2Loss of time
If traffic scheduling is adjusted based on predicted handover times, then traffic delivery timeliness is improved, but system complexity increases due to handover prediction processing
Solution Approach 1:
The system introduces handover prediction reports as an intermediary information element that mediates between UE handover status and RAN node scheduling decisions. These reports provide structured prediction information (including predicted handover time and confidence levels) that simplifies the scheduling logic, allowing the RAN node to make timely traffic delivery decisions without implementing complex handover prediction algorithms itself.
Solution Approach 2:
The RAN node determines pending traffic capable of being delivered before handover based on predicted handover times, performing preliminary traffic identification and scheduling actions. This advance planning allows the system to prioritize and deliver time-sensitive traffic before handover occurs, reducing delivery delay while using relatively simple scheduling rules based on the prediction reports.
3Measurement precision
If handover prediction reporting is configured with strict criteria, then prediction accuracy is improved, but signaling overhead increases
Solution Approach 1:
The system configures handover prediction reporting with optional criteria that can be selectively applied. The RAN node can request prediction reports with varying levels of detail and confidence thresholds based on current network conditions and QoS requirements. This partial application of strict criteria allows the system to obtain sufficient prediction accuracy for critical scenarios while avoiding excessive signaling overhead in normal operating conditions.
Data Source
AI summary
A user equipment may receive, from a serving radio network node, handover prediction configuration information indicative of at least one handover prediction criterion. A learning model executed by the user equipment may determine, based on the at least one handover prediction criterion, at least one handover event prediction and a confidence level corresponding to the learning model. The user equipment may adjust a time criterion of the at least one handover prediction criterion if the confidence level does not satisfy a confidence level criterion of the at least one handover prediction criterion. Based on the adjusted time handover prediction criterion, the learning model may determine an updated handover event prediction and an updated confidence level that satisfies the confidence level criterion and may report the updated prediction to the node. The node may adjust scheduling of delivery of traffic directed to the user equipment based on the updated prediction.