QoS Analytics Confidence Thresholding for 5G Prediction Notifications
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing 5G networks lack efficient mechanisms to provide confidence thresholds for quality of service (QoS) sustainability analytics, leading to inaccurate predictions and unnecessary resource consumption in notifications.
Innovation Solution
Implementing confidence thresholds in analytics requests or subscriptions, allowing consumers to set specific confidence levels for QoS metrics, thereby defining conditions for prediction events and notifications, reducing inaccuracies and conserving processing resources.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If confidence thresholds are implemented in analytics requests, then prediction accuracy and notification reliability are improved, but device complexity and processing overhead increase
Solution Approach 1:
The patent introduces confidence thresholds as a new parameter in analytics requests, allowing consumers to specify the minimum confidence level required for predictions to be considered valid. This parameter change enables the system to filter out low-confidence predictions, improving notification reliability without fundamentally altering the core analytics processing architecture.
Solution Approach 2:
The confidence threshold is evaluated preliminarily during the analytics request processing stage, before final predictions are generated and notifications are sent. This preliminary filtering action prevents unnecessary processing of low-confidence predictions, reducing overall system complexity while maintaining high reliability for critical notifications.
2Measurement precision
If confidence thresholds are enforced for all analytics predictions, then prediction accuracy is improved, but processing resources and computational overhead increase
Solution Approach 1:
The system dynamically adjusts the evaluation of confidence thresholds based on the analytics request parameters and consumer requirements. Rather than applying a fixed high threshold universally, the system adapts the confidence level requirements to match the specific needs of each analytics consumer, optimizing the balance between prediction accuracy and processing resource consumption.
Solution Approach 2:
By allowing consumers to specify their required confidence levels as parameters in analytics requests, the system can adjust the threshold requirements dynamically. This enables higher accuracy for critical predictions while allowing lower thresholds for less critical analytics, thereby reducing overall processing resource consumption.
3Reliability
If confidence thresholds are set high to ensure accurate predictions, then notification reliability is improved, but the frequency of notifications decreases
Solution Approach 1:
The patent enables consumers to configure their preferred confidence threshold levels as request parameters. This allows each consumer to set the appropriate balance between notification reliability and frequency based on their specific needs, rather than using a uniform threshold that would either overwhelm the system with false alarms or miss important predictions.
Solution Approach 2:
The system segments the analytics notification process into different confidence levels, allowing consumers to subscribe to different threshold requirements simultaneously. This segmentation enables critical consumers to receive only high-reliability notifications while less critical consumers receive more frequent updates, optimizing both reliability and productivity across the board.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Certain example embodiments provide systems, methods, apparatuses, and computer program products for providing confidence thresholds in an analytics request or subscription. For example, certain embodiments may include a confidence threshold for each reporting threshold (e.g., quality of service (QoS) metric-specific reporting threshold) in an analytics request or subscription (e.g., a network data analytics function (NWDAF) QoS sustainability analytics request or subscription). The confidence threshold (lower bound) can be determined by the consumer application based on their individual needs, e.g., based on the cost or impact of the compensating actions on the predicted QoS sustainability notification. The confidence threshold, in combination with the reporting threshold, may define the conditions for analytics events and/or notifications (e.g., QoS sustainability prediction analytics events and the related notifications).