Helix® Performance Management (PM) is an ultra-scalable system for managing the end-to end performance and QoS of complex telecom networks of all types. Its array of integrated software modules enable operational and engineering teams to meet and exceed their goals:
- Network planning and dimensioning: correctly prioritize network expansions and upgrades.
- Network operations: reduce the effort to identify performance issues, manage KPIs and counter sets, and process large volumes of data.
- Service operations: minimize service degradations and impact on end customers by enabling early detection of traffic patterns and resolution of low performing networks.
BIG DATA IN ACTION
How big is ‘Big Data’? When it comes to your network performance, huge amounts of data arrive from multiple domains, technologies, equipment vendors, service probes and external resources. Helix® PM collects and analyzes hundreds of millions of performance measurements per day and translates millions of data records, in diverse formats and in real-time. The solution supports ‘Big Data’ storage using Hadoop framework and traditional relational databases using Oracle. This dual support provides CSPs with the flexibility to align with both long-term data storage and day-to-day operational needs.
NEAR REAL-TIME KEY INDICATORS
Helix® PM provides strong data-aggregation capabilities for calculating Key Performance Indicators (KPIs) and Key Quality Indicators (KQIs), enhancing system performance and traffic-analysis calculations. For example: Peak and Busy Hours, as well as automatic ageing (thin-out) of operational measurements (OMs). The solution also generates Threshold Crossing Alarms (TCAs) based on a wide variety of conditions and supports multiple severity levels.
SERVICE PERFORMANCE MANAGEMENT
Helix PM helps CSPs to improve Quality of Services and monitor end-to-end services for mobile, fixed, IP and OTT services. In addition to performance management, the same large data processing capabilities and analytics algorithms can be applied to service-oriented KPIs and KQIs. As a result, service operations centers (SOCs) can quickly identify the Service Level Objectives (SLO) violation per service, as well as further investigate the underlying issues to understand the root cause. The available tools help them focus on problems with high impact on services and customers, diagnose their effect, and automate the resolution process.
DEEP INSIGHT INTO NETWORK PERFORMANCE
Helix® PM includes a set of analytics and machine learning algorithmic tools that provide an automated view of the network performance and its reliability. These tools improve the ability of engineering and operations teams to execute key responsibilities and daily tasks. A range of options is available, from threshold crossing monitoring, through adaptive and machine learning thresholds, advanced data forecasting and statistical anomalies detection. Using the analytics tools CSPs gain an insight into the nature of their network and user traffic, the patterns, and the behavior of different KPIs and KQIs, as well as their internal effect on each other and their impact on the tasks at hand.
TrafficGuard – ANALYSIS AND TROUBLESHOOTING
TrafficGuard monitors performance patterns for service degradations. It uses smart thresholds with several simple-to-complex comparison rules to proactively analyze evolving network patterns and create threshold crossing alarms (TCAs).
Adaptive Thresholds – AUTOMATICALLY LEARN YOUR NETWORK BEHAVIOR
Adaptive Thresholds automates the thresholds’ creation and maintenance. Built on top of TrafficGuard, it significantly reduces the number of required threshold rules in the system and the time to deploy new services. Adaptive Threshold uses several statistical algorithms that can calibrate and fine-tune the thresholds to detect non-trivial problems in an early stage.
Forecasting – PREDICT YOUR NETWORK BEHAVIOR
Forecasting enables users to schedule regular automatic executions of data forecasts and trend analysis for any counter, KPI or network element in the system. TEOCO’s patented, self-developed algorithm efficiently applies the SARIMA (Seasonal Auto-Regressive Integrated Moving Average) model on large datasets. The results of these calculations are stored in the database and can be used for reporting and alarms creation.
Anomaly Detection – UNCOVER IRREGULAR BEHAVIOR IN YOUR NETWORK
Anomaly Detection is a machine-learning algorithm that automatically identifies entity instances with high level of abnormal KPIs behavior or non-trivial KPI-to-KPI relation. The algorithm learns the behavior of the KPIs, compares the patterns, and automatically identifies anomalies. The score results are displayed per instance, enabling easy identification of the anomalies and their distance from the common cluster or the common behavioral patterns in the network.
FAST INVESTIGATION TOOL
Part of the Helix® Performance Management solution, Sentinel provides a centralized user experience and investigation tool for accessing all modules, views and user actions, including fast queries on large data sets. Built as an intuitive, process-driven interface with convenient widgets, Sentinel enables users to control the views and customize them to serve their immediate needs as they monitor the service levels, investigate network performance issues and quality degradation.