Radio Access Network Data Collection Cost Budgeting
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The extensive resource usage and increased load in data collection for machine learning tasks in radio access networks lead to excessive resource consumption, which is not efficiently managed, causing potential cost overruns and inefficiencies.
Innovation Solution
Implementing a system that determines and manages a cost budget for data collection, allowing network nodes to estimate and transmit cost values associated with data collection tasks, enabling controlled resource usage by prioritizing data requests based on allocated budgets and thresholds, thereby preventing excessive data collection loads.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If extensive data collection is performed to enable machine learning inference, then the quality and usefulness of ML results is improved, but resource consumption and network load increase significantly
Solution Approach 1:
The patent applies parameter changes by dynamically adjusting data collection parameters (such as data volume, collection frequency, and granularity) based on the cost budget. The system modifies these parameters to balance ML inference quality with resource consumption constraints, ensuring that data collection remains within acceptable cost limits while still providing sufficient data for meaningful ML results.
2Productivity
If more network data is collected for ML tasks, then the effectiveness of ML optimization is improved, but the load at data interfaces and power consumption at network entities increase
Solution Approach 1:
The system dynamically adjusts data collection parameters based on the cost budget to optimize the balance between ML effectiveness and power consumption. By modifying parameters such as data collection intensity and scope according to available resources, the system ensures that ML tasks receive sufficient data for effective optimization without causing excessive power consumption at network entities.
3Measurement precision
If comprehensive network information is gathered to support ML tasks, then the accuracy of ML-based network optimization is improved, but the cost and complexity of data collection infrastructure increases
Solution Approach 1:
The patent implements parameter changes by adjusting data collection characteristics (volume, detail level, collection methods) based on the cost budget. This allows the system to gather sufficient network information for accurate ML-based optimization while controlling the complexity of the data collection infrastructure by adapting collection parameters to resource constraints.
Data Source
AI summary
Example embodiments enable controlling consumption of additional resources required for performing data collection in a radio access network by estimating costs associated with various data collection tasks. The costs may be exposed to data consumers such that is possible to keep the additional resource consumption within cost budget(s) allocated for particular data consumers, network nodes, and/or network interfaces. Example embodiments enable preventing excessive data collection load, which may be for example caused by various machine learning tasks performed at the radio access network. Apparatuses, methods, and computer programs are disclosed.