RAN Capacity Forecasting for Nonlinear KPI Threshold Planning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional time series forecasting techniques fail to accurately forecast radio access network (RAN) capacity for non-linear key performance indicators (KPIs, leading to poor end user experience, service degradation, and resource inefficiencies due to under- or over-provisioning.
Innovation Solution
A forecasting system utilizing Prophet and XGBoost models, with feature engineering and rotational stitching, to predict RAN capacity thresholds and adjust accordingly, optimizing resource allocation and infrastructure improvements.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If traditional time series forecasting techniques are used for RAN capacity planning, then the system is simple and easy to implement, but the forecasting accuracy deteriorates for non-linear KPIs
Solution Approach 1:
The patent combines multiple forecasting models (Prophet for trend/seasonality, XGBoost for non-linear patterns) into a composite forecasting system. This ensemble approach leverages the strengths of each model to achieve superior forecasting accuracy for non-linear KPIs while maintaining manageable system complexity through modular architecture.
Solution Approach 2:
The forecasting process is segmented into distinct stages: data preprocessing, feature engineering, model selection based on seasonality metrics, and forecasting. This segmentation allows each component to be optimized independently, improving overall accuracy without proportionally increasing complexity.
2Loss of energy
If RAN capacity is under-provisioned to conserve resources, then resource efficiency improves, but service degradation occurs
Solution Approach 1:
The system continuously monitors actual RAN capacity utilization and compares it with forecasted values, using this feedback to refine future forecasts and adjust capacity provisioning decisions. This closed-loop approach ensures resources are allocated efficiently while maintaining service quality through continuous optimization.
Solution Approach 2:
The system performs capacity forecasting in advance to predict future RAN capacity requirements, enabling proactive resource allocation before service degradation occurs. This preliminary action allows network operators to plan capacity upgrades ahead of time based on accurate predictions rather than reactive measures.
3Reliability
If RAN capacity is over-provisioned to ensure service quality, then reliability improves, but resource efficiency deteriorates
Solution Approach 1:
The system dynamically adjusts capacity provisioning parameters based on forecasted demand patterns, seasonal metrics, and historical data. Instead of static over-provisioning, the system optimizes capacity parameters in real-time to match actual needs, maintaining service quality while eliminating unnecessary resource allocation.
4Quantity of substance
If traditional forecasting methods are used, then implementation cost is low, but forecasting accuracy for non-linear KPIs deteriorates
Solution Approach 1:
The system introduces intermediate processing steps including feature engineering and seasonality metric calculation that bridge the gap between simple input data and accurate forecasts for non-linear KPIs. These intermediary transformations enable accurate forecasting without requiring excessively complex or expensive models.
Data Source
AI summary
A device may receive load data identifying a load on a radio access network (RAN), and may select one or more time series forecasting models and a classification model based on seasonality metrics associated with the load data. The device may process the load data, with the one or more time series forecasting models, to forecast a capacity for the RAN, and may process the load data and the capacity, with the classification model, to determine whether the capacity exceeds a capacity threshold. The device may selectively determine that the RAN does not need an upgrade based on determining that the capacity fails to exceed the capacity threshold, or may adjust, based on determining that the capacity exceeds the capacity threshold, the capacity to generate an adjusted capacity. The device may perform one or more actions based on the adjusted capacity.