Unlocking the Potential of Telemetry in Modern Broadband Networks
Automation has become critical for enhancing efficiency in operating modern broadband networks in a world that is now incredibly dependent on various digital technologies. High-quality telemetry data is pivotal in harnessing automation’s full potential, enabling applications such as analytics, automated troubleshooting, and AI/ML to thrive.
Gone are the days when telemetry was seen as bulk data collection at regular intervals, typically every 15 minutes. Traditional Operations, Administration, and Maintenance (OAM) techniques and protocols are being surpassed by more advanced push-based streaming telemetry systems, offering enhanced flexibility, scalability, accuracy, coverage, and performance. This paradigm shift empowers network operators to embrace automated control, data-driven decision-making, and closed-loop automation.
Real-time, structured telemetry data is the key to unlocking various use cases, including network automation, traffic optimization, anomaly detection, automated troubleshooting, and preventive care.
Telemetry, the central nervous system of modern broadband networks, is the linchpin of intelligent, automated networks. Its role in remote data collection, processing, storage, and consumption is akin to the neural pathways in our bodies. Good telemetry captures data precisely, transmits it nearly instantaneously, and provides essential information for real-time monitoring, AI/ML insights, and closed-loop automation.
The advent of SDN and cloud-native architectures has propelled telemetry into a new era, enabling network operators to capture and process vast amounts of data with greater accuracy and flexibility. Standardization and data governance are essential, allowing telemetry to support a wide range of use cases, from network monitoring to marketing opportunities. In the ever-evolving landscape of modern broadband networks, telemetry is the vital component that ensures efficient and effective network operations.
Network Evolution to SDN and Cloud-Native Designs
The evolution of networks towards Software-Defined Networking (SDN) and cloud-native architectures has revolutionized how we view telemetry. SDN, especially in access networks, has ushered in an open, modular, disaggregated cloud-native architecture. This transformation involves decoupling traditional node architecture, transparent APIs, flexible YANG-based device modeling, and the independent operation of individual network components. Standardized interfaces for management, control, and orchestration, as defined by Broadband Forum’s CloudCO, have become instrumental in facilitating effective automation.
These standardized interfaces give operators a holistic view of the fixed access network across multi-vendor, multi-technology environments. With SDN controllers creating virtual networks in the software layer, the authoritative data source now resides in the cloud rather than individual network elements. This centralized data storage and management approach streamlines configuration updates and data access while enabling extensive analytics and machine learning capabilities.
Scaling Up the Level of Automation Incrementally
In cloud-native network architectures, the focus has shifted from nodes to controllers. Traditional node-centric data reporting, often limited to 5-15 minute intervals, hindered real-time monitoring and scalability. Polling data from nodes resulted in higher CPU loads and limited network management system (NMS) intelligence.
On the other hand, SDN controllers leverage push-based streaming telemetry to monitor a broader array of network devices and process more parameters, often up to 20 times more counters than legacy NMS solutions. These controllers offer real-time network insights and empower operators to optimize data collection models, define key performance indicators (KPIs), trigger network policies, initiate workflows, and engage in closed-loop optimization. As automation increases, it’s essential to allocate adequate resources to data lakes and telemetry systems, as the volume of network data can substantially impact system performance.
The Foundation for Automation: Data Governance
To achieve high-level automation, data governance becomes paramount. Data should be structured and readily accessible through open APIs, eliminating data acquisition and storage duplication. Standards like IETF RFC 9232, ETSI Zero Touch Network and Services Management (ZSM), Generic Autonomic Networking Architecture (GANA), and others play vital roles in shaping the future of network telemetry.
These standards emphasize high precision, centralization, consistency, standardization, scalability, ease of consumption, customizability, low latency, and security in telemetry systems.
Modern Network Data Telemetry Systems
Modern network data telemetry systems should possess the following attributes:
- High Precision: Capable of capturing complex data in the correct format, including configurations, logs, alarms, and counters.
- Centralized: Avoid fragmented views of data by centralizing storage.
- Consistent: Ensure high-quality data in real-time and historical contexts.
- Standardized: Utilize non-proprietary data formats for compatibility with machine learning.
- Scalable: Offer scalable database management in the cloud.
- Consumable: Facilitate easy access for external systems without data duplication.
- Customizable: Allow operators to define their own data models.
- Low Latency: Provide time-stamped data for immediate processing.
- Secure and Private: Comply with data protection regulations.
Modern Data Telemetry for Efficient Monitoring
In contrast to traditional monitoring platforms, SDN programmable networks leverage push-based streaming telemetry. Network elements push data, including statistics, alarms, state information, and more, to subscribed collectors based on predefined scopes and frequencies. This approach allows precise data selection, avoiding the transmission of unnecessary information. Telemetry frameworks support steady-state scanning and intensive collection modes, adapting to evolving application needs.
This flexibility is crucial for efficient monitoring and analysis.
Data Telemetry: The Central Nervous System of Networks
Imagine telemetry as a network’s central nervous system (CNS), akin to the human nervous system. As our CNS processes sensory input, telemetry collects, processes, and relays data from various network nodes. The efficiency of this process is similar to how our CNS captures sensory input with precision and transmits it almost instantly to the brain.
In traditional networks, data from network nodes is typically collected for 5 to 15 minutes, with most data aggregation and metric computations happening within the nodes. This approach is akin to a delayed or limited sensory response. However, data is actively pushed from nodes (push-based telemetry streaming) to a central data repository in the cloud in SDN-based networks. This paradigm shift allows SDN controllers to monitor many network devices and process many data points, sometimes up to 20 times more counters than legacy systems. Data can be pushed at high frequencies every 5 seconds, enabling real-time monitoring, AI/ML insights, and closed-loop automation.
Adaptive Telemetry: Mimicking Human Flexibility
One fascinating characteristic of telemetry is its flexibility, mirroring the human ability to focus on or suppress specific sensory signals. Adaptive telemetry techniques support dynamic adjustments to data collection based on changing application needs. This ensures efficient monitoring and analysis by accommodating steady-state scanning and intensive collection modes. Operators can select the amount, complexity, and frequency of the data they wish to work with, just as we can choose to pay attention to specific sensory inputs.
Telemetry and Standardization
In telemetry, standardization plays a crucial role in ensuring interoperability and compatibility. Machine learning algorithms thrive on standardized data formats, making comparing and analyzing information from different sources easier. Standards like BroadBand Forum TR-436 and ETSI Zero Touch Network and Services Management (ZSM) actively shape telemetry’s role in automated intelligent management and analytics. These standards emphasize the importance of structured, standardized, and non-proprietary data for a seamless flow of information.
Customer Use Cases
The true power of telemetry lies in its ability to unlock a wide range of use cases across network operations. Whether it’s network monitoring to understand performance, network assurance to preempt anomalies, or network engineering to meet performance targets, telemetry is at the heart of these endeavors. It also extends to customer care, where insights from telemetry can resolve customer issues more efficiently, and marketing, which can utilize network data for product launches and upsell campaigns.
- Network Monitoring: Understanding network and service performance.
- Network Assurance: Proactive remediation of network anomalies.
- Network Engineering: Measuring against design and performance targets.
- Network Planning: Design, upgrade, and dimensioning advice.
- Customer Care: Resolving customer trouble tickets.
- Marketing: Launching new products and improving upsell campaigns.
Download a Copy of Nokia’s White Paper on This Subject >>
About Infinity Technology Solutions
Infinity Technology Solutions specializes in broadband and critical communications infrastructure development. We help our channel partners create and deploy private wireless, microwave backhaul, IP/MPLS, and optical networking technologies.
For more information, give us a call or fill out the contact form below.
