Fierce's Cloud 101 Q&A series goes back to basics, aiming to help readers build in-depth knowledge about key cloud concepts and the overall market through progressive interviews with industry experts.
You can view the first installment in the series here.
This week's interview features Marisa Viveros, VP of strategy and offerings for the Telecom, Media and Entertainment Industry at IBM. The interview transcript has been lightly edited.
Fierce Telecom: Who can use the cloud? Is it just wireless operators, or can wireline operators and enterprises also tap these capabilities?
Marisa Viveros: Enterprises have been at the forefront of using the cloud, both private and public clouds. Wireless as well as wired operators are using the cloud for two aspects. One is for hosting all their business applications. For instance, user profiles, routing logs, billing data and all the information that a telecom company stores and maintains in a secure manner about their customers. Given the sensitivity of this data, many times telcos choose to host this data via a private cloud, but they’re using the principles of virtualization and cloud native to get value from this data.
The second way they’re using the cloud is to modernize their networks. When we think about network workloads – voice processing, video processing, content delivery or even the connection to a website – most of these functions are virtualized today. Again, most of that is currently in a private cloud, but telcos are considering migrating those functions into the public cloud. This is driving an increased need for a hybrid cloud approach that is flexible enough to support these workloads across different environments.
FT: What is the difference between the cloud and the edge, if there is one? From a telco perspective, what types of processing should be done in the cloud vs at the edge?
MV: Edge computing is a topic that is coming very much into fashion these days, and it is complementary to the cloud.
When we think about workloads, the core processing that is needed for a telco or enterprise runs in the core of the cloud, the central cloud. Now edge computing has emerged with the opportunity to process data closer to where that data is produced.
For example, think about going to a stadium where there are cameras being used for surveillance and safety. Not all that video needs to go to the central cloud to get processed. An edge computing platform can be at the stadium and processing can be done right there. It is about minimizing the data transfers you have to do from where you capture the data, so actions can be taken much sooner than was expected in the prior model. Some aggregated data can still be sent to the cloud for auditing purposes.
Edge computing is evolving as more sensors are applied across enterprise use cases, whether it is a city, hospital, a manufacturing plant, enabling all that data to be captured and processed right where it is produced for maximal benefit. These are the kinds of applications that also benefit from very low latency in the response time.
FT: From a telco perspective, what applications and services have gained the most traction thus far and why?
MV: There is kind of a maturity curve going on, given telcos were slower than other industries in adopting the cloud. They started by putting a lot of emphasis on private cloud. Many of the workloads or applications they are running are core processes – all of the human resources processing, supply chain management and services to their users.
Today, the new workloads that are being processed on the cloud are the data management, analytics and AI algorithms. One of the most popular key processes is real-time predictive maintenance. That's a typical process that is using AI capabilities and automation in the cloud, enabling flexibility to burst-out to the cloud, as needed.
So the telcos started with the core applications to run their business and then adopted micro services with most applications, and now they are moving into adopting AI and automation in order to perform real-time predictive maintenance for their own business, as well as services that can be offered to enterprises.
FT: How are cloud applications evolving in terms of architecture? Why is this important?
MV: Virtualization was the first wave of this architecture, which meant no longer depending on hardware but now depending on software architectures. The next wave transformed this architecture into microservices, breaking monolithic applications into smaller chunks that could connect via APIs. The third wave, being done today, is what we call “cloud native,” which are applications that are designed for the cloud, no longer thinking about hardware systems. It will take these several years to be fully adopted.
Why this evolution is important is because of application portability and system management simplification. For software upgrades, now you don’t have to wait for the next cycle to upgrade in two or three weeks. Now we can start doing upgrades many times during the day. That flexibility and the agility of development is what is driving this new architecture. It’s all about making more agile, more responsive systems. For physical infrastructure, telcos can leverage the frequent renewal of resources that happen on the cloud.
FT: Is there a connection between open RAN and the cloud?
MV: Open RAN is addressing the last section of the network architecture that needs to be virtualized, so you can use any radio, any software to bring it all together between the central unit and the distributed units and connect that to the central systems.
Can open RAN run on the cloud? Yes, it can run there, and what we also see is its migration to edge computing facilities. Those facilities now can host these open RAN and distributed networks in a much more efficient manner, and that’s leading many telcos to embrace an open hybrid cloud network architecture that supports this agility.
It is, again, migrating the power of the cloud into the network, and extending that power not only to the core network but also to the edge of the network, which is the RAN where the radios are. It is a development that is starting with pilots, and we will see more and more of these deployments as we move forward.
FW: What is currently the biggest barrier to cloud adoption? What can be done to help overcome this?
MV: Telcos have had a complex cloud journey. They’ve faced two major hurdles: security concerns and trying to adopt flexible cloud technologies while using traditional methodologies to manage the systems. Basically, telcos could not benefit from the cloud because their processes remained in the past; yet they experienced increasing volumes of data spread across different environments.
That's why we have encouraged our clients to adopt a hybrid and secured cloud architecture that enables deployment choice and flexibility and supports a diverse ecosystem of partners that can speed innovation. It's sort of a mindset and education process there, and also acquiring the skills to implement cloud operating models and methodologies.
The success of cloud in telecommunications depends on the adoption of hybrid cloud; learning from other industry verticals, wherein mission critical workloads have been migrated to an open and scalable cloud architecture that has security built in. Adopting the cloud operating model for on-going management will allow telcos to have the agility their clients expect.