Cloud-Native OSS for Telecom: 4 Reasons to Choose Containerized Deployments
5G connectivity has the potential to accelerate growth across adjacent industries — from automotive to retail — that have grown more “digital” and, therefore, reliant on cellular network services. For telcos, this rapidly increasing state of connectivity opens new B2B and B2C revenue streams.
5G Capabilities and Use Cases
However, to profit from the growing demand for better connectivity, the sector will have to complete an admirable scope of technical transformations across the OSS technology environment.
Traditionally, telecom has always been a “closed” industry. The entry bar was high because of the expensive (and limited) government licenses, as well as significant investments in infrastructure. Most telecom networks were built using the same standards and equipment, supplied by the same manufacturers. Respectively, the network quality and service levels became homogenized.
Progressively, the telecom sector evolved into an industry with little differentiation. Everyone was in the business of providing customers with communication capabilities — voice calls, texts, and broadband connectivity. There was little room to compete beyond pricing.
However, the rapid expansion of cloud technologies and subsequent “IT-ization” of telecommunication services gave players two new competitive strengths:
- Higher operational efficiency
- Better service levels
Cloud-based OSS architecture has allowed market leaders to dramatically reduce the internal operating costs and level up service quality through end-to-end network visibility, faster deals management, and rapid roll-out of new service packages, using virtualized infrastructure.
The upcoming arrival of 5G further magnifies the importance of speed and effective resource orchestration. Novel network design approaches such as software-defined networking (SDN) and network functions virtualization (NFV) enable access to an earlier impossible degree of:
- Rapid network scaling
- Effective resource orchestration
- Network automation
The upcoming surge in network traffic and dynamic user needs also puts greater pressure on legacy OSS/BSS systems that were never meant to support multi-play commercial models. Cloud technologies will play a focal role in this transformation.
Thanks to embedded elasticity, in-built redundancy, and enhanced collaboration capabilities, the cloud environment allows telecoms to rapidly respond to new market demands.
In a cloud-native environment, telecoms can build microservices-based applications and deploy them as containers. Then apply the following DevOps practices to improve the cadence and efficiency of new deployments:
- Continuous Integration and Continuous Delivery (CI/ CD)
- Continuous testing (CT)
- Resources orchestration
- Workflow automation
Such a transition to cloud-native, containerized OSS solutions carries a host of benefits. The British telecom player, Truphone, for example, managed to reduce the total cost of ownership of OSS applications by 60% after migrating to the public cloud.
The savings came from eliminating CAPEX spending on local infrastructure, as well as third-party software fees. Additionally, the Truphone managed to replace over 5,000 manual tests with an automated CI/CD pipeline that guaranteed better coverage and faster execution time. This has led to a further operational costs reduction.
Truphone is hardly the only telecom pursuing OSS cloud migration. In fact, over 90% of global CSPs plan to move at least some part of their OSS/BSS to cloud infrastructure by the end of 2022. Among them, over 60% are considering hybrid cloud deployments for OSS.
The rationale behind such unified interest is simple: the legacy OSS service layers telecom providers rely on can no longer cope with the new market demands:
If these are the challenges your company recognizes as well, here are four ways cloud technologies can help you resolve them.
B2B clients expect ultra-high network availability performance and low latency from 5G providers. Currently, networking issues are an unpleasant nuance for most clients. However, in the future, millisecond delays can lead to dire outcomes — an autonomous vehicle going off the highway or a remote medical procedure, halted by service unavailability.
To support the new latency targets of 1ms, many telecoms are considering network slicing — a method for digitally provisioning highly-available network segments to certain customers on demand. Grounded in network virtualization, network slicing scenarios will also require transformations in OSS/BSS.
In particular, telcos will need to create more customized OSS suites to rapidly provision the requested slices and assure the contract-set SLA targets. Cloud-native OSS enables faster services provisioning. Thanks to the orchestration and automation of manual tasks, such systems are capable of auto-configuring the cell sites to provide the requested coverage by the client. Similarly, the SLAs and other policies can be codified into your system to ensure automatic compliance at every step of the initiation and provisioning.
2. End-to-End Visibility
Telcos estates now feature a multi-vendor, multi-location collection of networking assets, both physical and virtualized ones. Managing such a rapidly expanding estate manually is no longer feasible.
A build-up of unprocessed raw log messages, alerts from edge devices, and other connected hardware makes it hard to discern the important performance and security events from day-to-day clutter.
To achieve end-to-end visibility into such complex environments, telecoms will need to consolidate reporting streams. Then establish better mechanisms for correlating data from different sources.
Cloud-based SIEM and SOAR solutions such as Microsoft Sentinel are of great help in this case since they can provide consolidated reporting. Important performance-related insights can be then sent over to the OSS module so that your teams could take rapid action to better coordinate the resources.
Alternatively, you can always opt to create more advanced performance management solutions. For example, Nokia and Vodafone recently launched a service assurance solution on the Google Cloud Platform. The new ML-powered anomaly detection system scans through Vodafone networks to locate network performance irregularities such as mobile site congestion, interference or unexpected latency. Then the system either applies self-healing scenarios or alerts the operating team to provision required capacities.
As mentioned earlier, cloud-native
technologies enable telecoms to launch microservices-based containerized
applications. A container hosts all the elements an application needs to be
rapidly deployed in any target environment.
Containerized OSS software applications have one major advantage: they are stateless, meaning they don’t host any data. Instead, they consume it via API from a separately hosted database.
The above means that containerized apps are easier to develop, debug, and re-deploy if needed. For example, if you decide to change the network function at some point, you will not need to worry about complex data migration. Container apps are also less complex and costly to maintain than VM-based applications, which are more dependent on the surrounding infrastructure.
These technical characteristics of containers, paired with DevOps principles of continuous deployment and integration enable telcos to rapidly launch new customer services and upgrade existing ones with new functionality.
Take it from Huawei. The company is in the process of migrating some 800 VM-based apps to the Kubernetes container platform. Up to date, they have moved around 30% of their estate, but are already seeing major improvements in their deployment speed — from one week to several minutes on average. Additionally, the team recorded a 20-30% reduction in operating costs.
Last, but not least, cloud platforms come pre-equipped with an attractive range of ML and DL capabilities, which telecoms can use to mine more valuable insights and deploy innovative customer-facing applications.
Taiwan Mobile and Nokia recently collaborated on an OSS framework with embedded AI capabilities. Nokia provided Taiwan Mobile with a cloud-native end-to-end network and service assurance solutions for monitoring network issues and reinforcing their impact on customer experience.
The first step in this project was to develop a centralized OSS-based data lake to enable better end-to-end visibility. Next, Nokia selected suitable data sources to establish analytics for better network resource quality and support other apps for improved fault management, configuration management, and performance management.
In the next stage, Nokia helped Taiwan Mobile integrate and customize its AVA Telco AI ecosystem — a cloud-based AI as a service platform, offering access to telco-specific full life-cycle management of AI/ML-driven applications. Using the provided out-of-box tools, Taiwan Mobile managed to deploy a host of innovative customer-facing products that led to impressive results:
What Stalls Public Cloud Adoption Among Telecoms?
Cloud technologies, and public clouds, in particular, enable a plethora of OSS management advantages. Still, the TM Forum’s research suggests that most telecoms have less than 5% of their operating software hosted in public clouds. Why is this so?
Several types of concerns surround
cloud-native OSS deployment:
- Migration complexity: Even the most basic deployment scenarios such as converting parts of monolith legacy apps to containerized, cloud-native solutions require a high degree of tactical planning to prevent service disruptions. Soundly, many cloud companies address this reluctance by creating telecom-specific migration offerings such as Azure for Operators. TM Forum has also launched Open Digital Architecture (ODA) — a reference architecture telecoms can use for cloud modernization projects. Finally, if you lack in-house resources, you can always consult with a telecom technology partner to receive technical guidance and proactive assistance with the execution of a project.
- Extra costs: Cloud modernization projects require a significant initial investment, but many telecoms already have the bulk of their IT budgets tied up in maintenance and hardware investment. In this case, it is important to remember that in the long-term perspective OPEX costs always end up being lower than CAPEX. Virtualized network architecture and supporting services are more affordable to maintain than on-premises hardware. Moreover, both also bring in new revenue streams and hence, proven ROI.
- Compliance and security: Some telecom remain concerned with the regulations, surrounding customer data storage requirements. Specifically, that most sensitive data can be hosted only on-premises. Indeed, this is often the case. However, regulations do permit processing anonymized customer data in the clouds, as well as other types of insights. Security-wise, public cloud services providers have made rampant progress in the number and sophistication of security controls available. Standard security network functions are natively available for all major public cloud providers. Additionally, you can add custom controls and purchase extra security tools for SOC.
At the end of the day, telecoms are in the business of providing a stellar customer experience. Given that consumers today have a growing array of choices — from OTT players eroding cable services and communications to digitalized competitors, offering more innovative price packages — telecoms are under pressure to level up their services. This improvement is impossible to pull off without the further transformation of the OSS/BSS stack and the advantages cloud technologies provide.
Likewise, regulations around data storage and
data privacy keep evolving to support secure, anonymized processing in the
clouds. While you may still need a local data center for the most sensitive data,
ignoring the cloud-enabled possibilities for growth can leave you at a
Contact the Infopulse telecom team to further discuss the opportunities cloud technologies bring in for the sector and different adoption scenarios.