Spotlight on Tech

Complex telecom ecosystems: Observability is the ‘X Factor’ for future telcos

By
Sree Nandan Atur
Technical Director – Symcloud
Rakuten Symphony
March 21, 2024
12
minute read

The telecom industry is at a critical juncture. We are transitioning towards a paradigm that's both technologically advanced and transformative. This shift demands a new focal point: observability. In the context of complex systems, observability isn't merely about monitoring systems; it's about understanding and controlling them. Rakuten Symphony's Telco Cloud solution stands out in this regard as a market-leading cloud-native Kubernetes/Orchestration solution, offering heightened control and insight into complex and dynamic telco ecosystems.

The past decade has been a period of unprecedented change in the telecom sector. Significant technological shifts and strategic realignments have contributed to these changes. When it comes to the quest to deliver new and innovative services to customers, the need for more flexible, scalable, on-demand deployment and cost-effective networks becomes non-negotiable.

Over a period of time, we saw the telecom sector evolve from a rigid, hardware-centric industry to a more flexible, software-driven one. The adoption of technologies like Software-Defined Networking (SDN), Network Functions Virtualization (NFV), Open RAN and cloud-native principles, are enabling the creation of networks that are more adaptable, scalable and capable of handling the demands of modern applications and services. As we look to the future, these trends are expected to continue, with further advancements in 5G, edge computing and network security driving the next wave of transformative change.

Here's a closer look at some of these trends:

  1. Software-hardware disaggregation: Traditionally, telecom networks were built on proprietary, monolithic hardware, where the hardware and software components were tightly integrated. Disaggregation has involved breaking down these traditional network components into smaller, interchangeable parts. This shift allows telecom operators to mix and match hardware and software from different vendors, leading to increased competition, lower costs and greater flexibility in network deployment and management. However, this also introduces complexities in orchestration and integration, demanding more sophisticated observability tools which can provide the real-time or near-real-time data from a complex and disparate system. Such a system involves legacy, cloud-based, edge, core, network and applications layers providing insights for meeting user demands in a cost optimized manner.
  2. Softwarization and virtualization: The industry has seen a significant shift towards the softwarization of network functions, moving away from hardware-centric models. This transition includes the adoption of SDN and NFV. SDN separates the network's control plane from the data plane, enabling centralized and flexible software-driven network management and automation. NFV replaces traditional network appliances with virtualized instances running on commodity servers, leading to improved scalability, cost of ownership and resource utilization.
  3. Open RAN (Radio Access Network): Open RAN disaggregates the RAN components of cellular networks, allowing components from different vendors to interoperate seamlessly. This open architecture fosters innovation, reduces costs and even supports features like massive MIMO, essential for 5G networks.
  4. Cloud-native: Cloud-native networks leverage the cloud's elasticity, resilience,and agility, using technologies like containers, microservices and hyperautomation through continuous integration/continuous deployment (CI/CD) pipelines. This approach allows telecom operators to deploy and scale applications more dynamically, improve time to market for new services, and enhance the overall network reliability and efficiency.
  5. 5G, IoT and the Edge: Arguably the most publicized development in telecom, 5G, brings faster speeds, high and reliable bandwidth, lower latency and the capacity to connect a vast number of devices. Many use cases today, such as IoT, autonomous vehicles and smart cities are reliant on 5G for assured performance. Edge computing – another key trend that implies moving workloads from centralized data centers to the edge of the network – is also critical for applications that require real-time processing. However, the emergence of 5G and Edge means telecom systems must manage and secure an ever-growing number of connections, each with varying service requirements regarding bandwidth, latency and mobility, often calling for rapid deployment and automated lifecycle management.
  6. The shift of telco workloads to the public/private cloud: In a way, this shift was inevitable. Cloud allows telecom operators to stay competitive, dynamically scale resources, reduce capital expenditures, and rapidly deploy innovative services among others. However, it also introduces challenges such as ensuring security, managing multi-cloud environments and integrating with legacy systems.

Observability is a challenge

Observability and control, particularly in cloud environments, present several challenges:

  • Dynamic and distributed nature: In the telecom sector, cloud ecosystems are inherently complex, leading to difficulties in monitoring and tracking the system's state and understanding the interdependencies between various components end to end.
  • Multitenancy and isolation: Ensuring security and compliance in multi-tenant architectures, integrating with diverse legacy systems and managing the performance-cost trade-offs further complicate the set-up.  
  • Multi-vendor, multi-cloud ecosystems: Ensuring interoperability, maintaining consistent performance standards and securing disparate systems pose significant challenges.
  • Performance and cost trade-offs: Gaining 360-degree visibility without adverse impact on system performance and cost is a continual challenge, especially in large-scale deployments.
  • Security and compliance: Cloud infrastructure and data might be spread across different regions; hence observability tools must correlate disparate features and address security challenges in open environments without violating regulatory requirements.
  • Evolving technologies and standards: Best practices for observability and control are continually evolving. Keeping up with these changes, integrating new tools and ensuring IT staff are up to date can be challenging.
  • Volume, variety and velocity of data: The sheer quantity and variety (logs, metrics, traces) of data generated, alongside the need for collecting, processing, correlating and analyzing it in real time place tremendous strain on traditional monitoring tools. This requires major investments in AI/ML technologies and supporting infrastructure.

Addressing these challenges requires a robust strategy combining advanced technological solutions, skilled personnel and continuously improving processes. As cloud environments become more prevalent in the telecom sector, developing effective observability and control mechanisms will be crucial for ensuring the reliability, performance and security of these complex systems.

The three pillars of observability

Observability in telecom transcends traditional monitoring. It offers a panoramic view of the network's health, performance, and security status. It's about predicting issues before they become critical and responding in real-time to dynamic network conditions.

It is underpinned by three pillars:

  • Enhanced data comprehension: It is important to consider the variety of data sources at play generated by physical and virtual network components in complex, modern-day telecom networks. This includes infrastructure, clusters, apps, and the network itself, each presenting data in various formats and standards. This data, consisting of both real-time and historical insights, must be processed and stored efficiently, while also meeting data ownership and privacy requirements. Observability delves into this data and provides a level of understanding, correlating the data obtained from all sources that is crucial for identifying patterns, predicting capacity needs, and spotting potential issues before they escalate.
  • Insightful predictive analytics: Observability goes beyond traditional monitoring, offering predictive insights through AI/ML that preempt network failures and congestion. The proven framework for network management, FCAPS (Fault, Configuration, Accounting, Performance, and Security) is closely related to observability. FCAPS provides a structured approach to monitoring and managing the network by observing parameters such as fault, availability, configuration changes, resource consumption, and security. It constitutes a holistic approach to enhancing network health and efficiency.
  • Proactive and intelligent actions that lead to monetization: This involves the use of a proactive approach, which means that networks aren't just reacting to issues but are continuously learning and improving. Prevention is not the only goal here. Proactive actions can lead to enhanced service offerings, better customer experiences and responsive services. This shift from viewing the network as a cost center to a source of value opens competitive differentiation by capitalizing on revenue-generating opportunities.

Open source is essential, yet not a standalone solution

Open source and Kubernetes have become the go-to platforms for managing cloud-native applications and 5G infrastructures. Kubernetes, with its closed-loop architecture, guarantees continuous operation and optimal resource allocation, enhancing scalability, flexibility and efficiency in multi and hybrid cloud environments. Its robust container orchestration capabilities allow operators to manage and scale services on demand, optimizing utilization and availability. Open telemetry enables operators to detect issues early, understand system dependencies and make informed decisions to improve performance and reliability. In distributed 5G and Edge ecosystems, open-source tools such as Prometheus, Thanos and InfluxDB, along with processing frameworks like Hadoop, facilitate the efficient collection, storage and analysis of multi-source, multi-format data across OSS systems.  

Kubernetes’ support for microservices-based architectures enables agility and rapid deployment, while features like self-healing, automated rollbacks and load balancing significantly improve system reliability and fault tolerance. However, it has its limitations.  

The shift to multi-cloud, multi-vendor ecosystems introduces complexity within environments marked by virtualization and disaggregation. The cloud's ability to swiftly scale resources up or down demands that observability control loops operate on the order of milliseconds. Pinpointing the precise dataset among massive data pools and routing it to the appropriate node for accurate observation is a formidable challenge, as are legacy integration, managing large-scale, distributed clusters and synchronization of diverse configurations and workloads. Issues such as interoperability and the security vulnerabilities associated with open APIs further complicate this landscape.  

Historically, the telecom sector's reliance on closed, proprietary data formats has impeded integration across multi-vendor ecosystems, complicating the comprehensive analysis, collection and storage of vital information. Often, achieving holistic observability requires reliance on third-party, vendor-agnostic tools, with the hope that the underlying systems are sufficiently open and compatible to allow for effective monitoring and management.

Mobility management as a use case

Consider a typical mobility management scenario in a 5G network. Radio Units (RU) absorb radio waves, transforming them into electrical signals transmitted via front-haul networks to Distributed Units (DU) at far-edge clusters. DUs, often requiring fast data processing, use accelerators and pass information through mid-haul networks to Central Units (CU), then to the core network and ultimately the internet. This process involves a mix of on-premises and cloud components, multiple vendors and various domains like RAN, core and application. Creating visibility and managing this process, especially in real-time and at scale, is an immense challenge.

Rakuten's Telco Cloud addresses the intricacies of modern telecom systems with nuanced control by providing a unified view across different layers and components, from the core to the edge, from physical infrastructures to virtualized functions.

  • Unified management across disaggregated layers: Rakuten Symphony's Telco Cloud provides a cohesive platform to manage and orchestrate the diverse and disaggregated components of 5G networks, from RU/DU/CU to core networks. It simplifies the complex web of hardware, software and cloud elements, ensuring they operate cohesively.
  • Enhanced observability and control:
    With its advanced observability features, Rakuten Symphony Telco Cloud offers real-time insights into every layer of the network. It can monitor the health, performance and security of each component, regardless of whether it's on-premises or in the cloud, proprietary or open-source. This holistic view is crucial for preemptive maintenance and swift issue resolution.
  • Streamlining operations with AI and automation:
    Leveraging AI and automated workflows, Rakuten Symphony Telco Cloud can predict potential issues, optimize resource allocation and ensure seamless handovers and load balancing. This not only reduces the operational burden but also enhances the network's efficiency and reliability.
  • Vendor agnostic integration:
    Recognizing the multi-vendor nature of modern telecom networks, Rakuten Symphony Telco Cloud is designed to be vendor-agnostic, offering seamless integration and management across different components and suppliers. This ensures that operators aren't locked into specific vendors and can choose the best solutions for their needs.
  • Facilitating cloud and legacy systems coexistence:
    Rakuten Symphony Telco Cloud bridges the gap between traditional legacy systems and modern cloud infrastructures, offering tools and interfaces to manage both effectively. This eases the transition to cloud-native setups and ensures continued performance and security across the hybrid landscape.

As we look to the future of telecom, observability emerges not just as a tool but as a strategic asset. Rakuten Symphony's Telco Cloud exemplifies this shift, offering a solution that is not just about monitoring and reacting but about understanding and proactively shaping telecom ecosystems.

Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.
Cloud-Native
Edge Cloud
Runtime Observability

Managing a multi-cloud environment is a significant challenge, particularly from the perspective of observability and data-driven analytics. Rakuten Symphony’s Sree Nandan Atur and industry experts recently discussed the key aspects of multi-cloud transformation at the Silverlinings Multi-Cloud Summit, with focus on observability. Watch the recording.

How can Symphony help?
Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.
By clicking “Accept All Cookies”, you agree to the storing of cookies on your device to enhance site navigation, analyze site usage, and assist in our marketing efforts. View our Privacy Notice for more information.