Unleashing the Power of Cloud: Automated Digital Operations

mavenir-network-cloud-native-software-web_1369278807

Since inception over 15 years ago, Mavenir’s foundations as a software company have been central to our mission and success. As telco operators have increasingly begun to deploy 5G standalone networks for delivering new 5G services such as network slicing, our expertise and experience in software development, deployment and optimization on COTS hardware has served as a key differentiator.   

Over the past five to six years, we have seen a significant increase in interest for hyper-scaled public cloud engagements – from telco operator hybrid cloud deployments to disaster recovery deployments. Our focus has been on ensuring that our software is able to run effectively in multiple environments, both in the private and public cloud environments. We took our continuous optimization mindset into delivering the very first 5G Open RAN-based deployment with DISH Wireless on AWS, and it has continued to drive us towards optimizing our products and moving beyond the proof-of-concept stage, into large scale deployments, ahead of most of the industry. 

Delivering the vision: one network, any cloud, all software 

A major focus of our product development strategy has been to validate that telco workloads can run optimally in any public cloud. From the outset, our approach has been to develop these capabilities into fully productized solutions, with the ability to run solutions in a public cloud. For Mavenir, it’s also about delivering a truly multi-generational technology, working across 2G, 3G, 4G, and 5G technologies, all in software. It’s at the core of our cloud-native vision: one network, any cloud, all software. 

To achieve this ambition, we continuously invest in our products to ensure that they deliver the optimal performance and integrate them natively into the tools that are already available in hyper-scaled environments. Eliminating the need for underlying third-party infrastructure is crucial for lowering costs. Hence, our approach is to deploy natively on Amazon Elastic Kubernetes Service (Amazon EKS). For DISH Wireless as a greenfield operator, to take an example, the ability to harness the CI/CD (continuous innovation/continuous deployment) pipeline when integrating with EKS has been critical, enabling them to leverage all the benefits of automation. 

This comprehensive integration approach has helped us rapidly accelerate development and demonstrate our product readiness in true GA quality across our full portfolio – including RAN, packet core, IMS, Messaging, Voice, and other solutions running in a public cloud environment.  

Following our successful engagement and landmark deployment with DISH Wireless, over the last three to four years we’ve rolled out this solution out to multiple customers. We are now seeing a growing demand for our solutions in Europe, including those from Telefónica to deploy full IMS. Meeting emerging market requirements has guided Mavenir in its product development and evolution, for example, ensuring ability of software to run on AWS Graviton processors in an optimized manner. This marks a major development since earlier, everything was Intel x86 instance-based. We’ve made a significant investment into making sure that software stacks from third parties which work well and are optimized on Intel x86 instances are fully optimized to leverage the Graviton feature set. Our agility and readiness in this area has further helped differentiate us in the market against some of our competition who are still at the proof-of-concept stage. 

Our journey with DISH Wireless: key learnings from a world-first deployment 

When we began our deployment journey with DISH, there were valuable learnings on both sides. AWS brought the vast experience of having automated millions of workloads for enterprises.  However, this differed from the way telecom workloads typically operate in terms of management and the requirement for very high availability. Our development journey served to educate both teams in terms of fully understanding how the CI/CD pipeline needs to adapt to support the unique requirements of telecoms. 

As one example, once everything is wired in, the specific workloads on the RAN side for the CU (centralized unit) now enable us to set up ‘00s sites/CUs, in a day. That level of productivity was simply not possible before we achieved cloud-native deployment.  

Automation has been another major advancement. In the past, the ratio of the number of employees needed to manage a set number of nodes was an order of magnitude compared to how this can now be managed within AWS. Moving from managing thousands of servers, each requiring the support of two or three people, to an automated process creates major efficiencies for an operator. 

Today, evolving core teams from operator-managed to automated environments represents the next challenge, including in brownfield deployments. Existing teams need to be brought in sync with the new streamlined methods of operation, where before they might have been used to scheduling a window at night to perform tasks, for instance. In this new mode, teams need to continuously develop, optimize, re-learn, and deploy fast. We’re going through that curve now with brownfield early adopters, and we absolutely believe that automation is the only way to realize the efficiency optimizations that are needed. 

Intelligent platforms: the path to 5G monetization

Historically, operators have invested billions of dollars into network capabilities, but with applications running over the top utilizing the underlying infrastructure without any constraints, the network is predominantly seen as just a connectivity pipe. 

The path to converting the network from functioning simply as a connectivity pipe into a fully-fledged services platform is by exposing Service APIs for Operational services and for Enterprises or Consumers. This means that the underlying products which enable the APIs must be cloud-native and need to be architected as microservices using Open APIs to allow full leverage. The telecom industry has had a long history of trying to monetize its networks through API exposure, but with limited success due to the complexity of the APIs themselves and absence of a platform which provides open, global, and accessible API solution with access to operator capabilities, in whichever network customers are in.   

Now there are new industry initiatives – such as the Catalyst Program and CAMARA Project – where key players are simplifying APIs hiding all the telco complexities, making it much more developer friendly and providing an API which run consistently between telco networks and different countries. Earlier this year, Deutsche Telecom and T-Mobile US announced the development of a platform with joint Application Programming Interfaces (APIs) – T-DevEdge – to make it easier for the global developer community to create new, connected solutions. Telefónica and DISH Wireless also have their own play in terms of network exposure, with many operators now getting on board. 

Some of the initial use cases from these initiatives are exploring how to achieve quality on demand (QoD). QoD API allows setting quality for a mobile connection in terms of guaranteed bandwidth, latency, jitter, and bitrate. This could be used by influencers or consumer journalists to publish their content even when in a congested area such as a stadium hosting major sporting events or a busy intersection. Another example of Service API implementation is Device Location, which provides the ability to check the location of the device. This enables targeted advertisement, geofencing or location-specific services such as in-stadium features for enhanced viewing of live sports matches. 

GSMA is defining IMS Data Channel which allows enriching native calling clients with interactive features using access to multiple OTT applications. IMS Data Channel is based on WebRTC’s Data Channel and is highly flexible and can carry any type of information such as texts or multimedia files or AR rendering information, etc. This introduces much-needed innovation to voice-based communication services. The Data Channel Services are exposed to third-party applications via Open APIs which will be used to deliver application content for richer use experience. Some of the use cases for IMS Data Channel are – Interactive calling which extends voice calling with screen sharing, co-browsing, co-editing, in-call AR/VR, social gaming and lots more. 

In addition to Service APIs, the Network-as-a-Service platform must also provide a common framework or APIs based utilities to enable an open and seamless way to consume and manage these APIs. Industry verticals that will soon be key consumers of these Service APIs will need a set of common services to simplify the implementation and deployment of vertical systems at large scale. 

Standards bodies have realized the need for such a common framework for both consumers and industry verticals. Two such frameworks are Common API Framework (CAPIF) and Service Enabler Architecture Layer for Verticals (SEAL), which provide common platform capability in the form of northbound APIs for application developers to implement real-time applications for consumers and industry verticals.

The Network-as-a-Service platform will also need to support Operations Service API in addition to Service API. Operations Service API will enable Orchestration and management of the Services exposing these Northbound Service APIs. These APIs will be critical for network automation which will support functions such as Instantiation, Auto-scaling, Auto-Healing, and Monitoring which enable network agility, resiliency, and scalability of this global services platform. 

Given the significant global investment that has been made in 5G technology, it is vital that the industry innovates ways to successfully monetize 5G and generate a return on investment. To achieve this, the ability to provide a platform which can be used to launch new services quickly becomes essential, and that will require embracing automation. Automation will enable launching of network functions to meet the requisite network performance. We are confident that the work Mavenir is doing together with our partners will play a pivotal role in moving the industry forward and empowering operators to realize the possibilities of full 5G monetization. 

Monetization Open RAN
Dark Mode