Edge Load Analytics via Application Data Analytics Enabler
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current 3GPP technologies lack support for edge load analytics, specifically in providing statistics and predictions for edge performance, failures, and service availability, and do not address data collection from various domains or utilize analytics for recommending actions to prevent overload in edge networks.
Innovation Solution
A computer-implemented method and apparatus for edge load analytics that collects data from multiple domains, processes it to provide statistics and predictions for edge node loads, and triggers proactive actions to manage overload scenarios by migrating services between edge and cloud environments.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If data analytics services are provided by NWDAF in 5G Core network, then network data analytics services can be supported, but edge load analytics and application-specific statistics cannot be provided
Solution Approach 1:
The patent segments the analytics service into two distinct components: NWDAF for network-level analytics and ADAES for edge/application-level analytics. This segmentation allows each component to specialize in its specific domain, with ADAES specifically handling edge node load analytics, service availability, and application-specific statistics that NWDAF cannot provide.
Solution Approach 2:
The ADAES acts as an intermediary layer between the 5G Core network and edge applications. It collects data from multiple sources including NWDAF, application servers, and edge nodes, processes this data to generate edge load analytics, and provides actionable insights to both network operators and application services, thereby bridging the gap between network-level and application-level analytics.
2Productivity
If edge analytics are implemented to provide statistics and predictions, then edge service optimization is improved, but system complexity increases
Solution Approach 1:
The ADAES is designed as a universal analytics service that handles multiple types of analytics requests through a standardized interface. It can process subscription requests for different edge nodes, generate various types of statistics and predictions, and serve multiple applications simultaneously, thereby reducing overall system complexity through consolidation rather than creating separate analytics systems for each use case.
Solution Approach 2:
The system manages complexity by dynamically adjusting analytics parameters such as data collection frequency, prediction time horizons, and aggregation levels based on service requirements. The ADAES can modify these parameters to balance analytics accuracy with system resource consumption, providing optimized performance without requiring complex manual configuration.
3Reliability
If proactive actions are triggered based on load analytics, then overload prevention is improved, but response time requirements increase system demands
Solution Approach 1:
The ADAES implements preliminary action by continuously monitoring edge node load metrics and generating predictions about future load conditions. When potential overload scenarios are detected, the system triggers proactive actions in advance, such as load balancing adjustments or resource allocation changes, before the actual overload occurs, thereby preventing service degradation while maintaining appropriate response times.
Solution Approach 2:
The system establishes a feedback loop where load analytics results are continuously fed back to the ADAES, which adjusts its predictions and triggers appropriate actions. This feedback mechanism enables the system to learn from actual system responses and refine its overload prevention strategies, improving reliability while optimizing response time based on real-time system conditions.
Data Source
AI summary
The invention provides a functionality for providing edge load analytics at an edge analytics producer as well as functionality for utilizing this edge load analytics for optimizing edge service performance. An edge analytics producer is configured to collect data and to perform edge load analytics considering data producers from different domains, to allow for edge analytics enablement. Further, based on the derived edge load analytics by the edge analytics producer, an analytics consumer is configured to generate a trigger event that indicates a predicted overload and a specific action to be performed.