Telecom Site Anomaly Detection for Capacity Planning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Telecommunications networks face challenges in efficiently understanding and predicting congestion, identifying increased traffic needs, and detecting abnormal customer behavior, leading to inefficient capacity planning and deployment of sub-optimum solutions.
Innovation Solution
A telecommunication site anomaly detection system that monitors and analyzes site behavior data to determine lower and upper limits for key performance indicators, detects anomalies, and recommends optimal capacity planning solutions such as temporary cell installations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional capacity planning methods are used, then network coverage is maintained, but congestion detection accuracy and capacity planning efficiency deteriorate
Solution Approach 1:
The patent replaces traditional manual capacity planning methods with an automated machine learning-based anomaly detection system. The system uses supervised learning models to automatically analyze network traffic patterns, detect congestion anomalies, and generate capacity planning recommendations, substituting human expert analysis with automated computational processes that provide both high accuracy and efficiency.
Solution Approach 2:
The system enables self-service capacity planning by automatically monitoring network sites, detecting anomalies without human intervention, and generating actionable insights. The anomaly detection system continuously analyzes site behavior data and autonomously identifies congestion patterns, allowing the network operator to benefit from continuous automated monitoring without requiring constant human oversight.
2Measurement precision
If more monitoring data is collected to improve anomaly detection accuracy, then detection precision improves, but system complexity and data processing requirements worsen
Solution Approach 1:
The patent extracts and focuses on specific key performance indicators (KPIs) that are most relevant for detecting capacity planning anomalies, rather than analyzing all possible network parameters. By selecting and monitoring only the most critical metrics such as site traffic, active users, and resource block utilization, the system achieves high detection accuracy while maintaining manageable complexity.
Solution Approach 2:
The system performs preliminary data processing and feature extraction to transform raw monitoring data into meaningful anomaly detection inputs. By pre-processing the data and extracting relevant features before analysis, the system reduces the complexity of the detection algorithm while maintaining or improving detection precision.
Data Source
AI summary
Systems and methods to detect abnormal behavior of cell sites and/or customers are disclosed. By detecting cell site and/or customer behavior anomalies, the system enhances capacity planning by helping understand congestion, more efficiently planning event sites, suggesting installation of temporary solutions, identifying when true traffic needs are increased, and detecting abnormal customer behavior and/or demand. The system accesses historical data for a set of KPIs (e.g., 3 months of hourly data that captures traffic, users, Physical Resource Block (PRB), and throughput). The system computes, for periodic time intervals (e.g., for each hour and each day), upper and lower limits for each site and for each KPI. Using this information, the system detects anomalies for current KPI measurements. The system sends alerts when anomalies are detected for a threshold period of time (e.g., when the KPI measurement falls outside of the computed upper and/or lower bounds continuously for 2-3 hours).