Spotlight on AI and Analytics | Disruptive Telecoms
“The scale and complexity of 5G networks demand that operators leverage AI/ML and analytics to take human operators out of the loop for time-critical decisions and actions…to drive closed-loop automation for network operations and service orchestration.”
With 5G fast becoming mainstream for telecoms today, it brings a greater level of complexity, more connected devices and more data running over the network. This increased level of complexity can only be handled with the help of AI/ML and analytics – and starting with the right architecture upfront to enable efficient growth and automation.
With its deep expertise and focus on telecom AI and analytics – Guavus is enabling its operator customers to gain new insights on their subscribers and network operations in real-time – and to improve their customer experience, reduce operations costs, and deliver innovative new 5G services.
Chris Neisinger, CTO of Guavus, speaks with Zia Askari from TelecomDrive.com about the top trends and issues Guavus is seeing working with operators globally, how operators are looking to better monetize their 5G investments, the role the 3GPP 5G Network Data Analytics Function standard is playing, and the latest edge-to-core-to-cloud analytics innovations from Guavus and what it has in store for the future.
A lot of operators are moving towards 5G today but they struggle to achieve better ROI projections. In such a scenario, how can they better monetize their network infrastructure?
We’re seeing a number of our operator customers looking to private networks, to Multi-access Edge Computing (MEC), and to new enterprise vertical services to better monetize their 5G infrastructure.
In terms of private networks, they’re looking to offer their enterprise customers individual slices from their 5G network. These slices can take advantage of our AI-based analytics for correlated customer and network operations insights, and provide advanced security.
With MEC, they can distribute compute and storage resources to the edge of their 5G network to assure the performance of new real-time 5G and IoT services. And by adding streaming analytics at the edge, they can further enhance performance.
And to continue to move up the value stack, many of our operator customers are eager to deliver new vertical services customized for their customers in key industries. By capturing analytics insights from their own mobile network data, combining that with their vertical customer data, they can then package and sell analytics as a new service to their customers in manufacturing, transportation, healthcare, local government, etc.
How can AI and the right analytics help operators to future-proof their investments?
One of the best ways to future-proof a new network investment is to get the architecture “right” upfront. Starting out with a strong architecture enables efficient growth and automation.
The 5G Network Data Analytics Network Function (NWDAF) architecture defined by 3GPP is a great example of this. Simple rollouts of 5G can be done without any automation by using brute-force rules, but taking this architecture “shortcut” will quickly bite the unprepared.
The scale and complexity of 5G networks demand that operators leverage AI/ML and analytics to take human operators out of the loop for time-critical decisions and actions…to drive closed-loop automation for network operations and service orchestration. NWDAF provides the AI/ML in the 5G Core and is a must for 5G automation.
Analytics can’t be a bolt-on later as operators might have been done in previous 3G and 4G generations. In defining the NWDAF standard — 3GPP, and the operators who contributed to the standard, have recognized that the complexity and scale of 5G demand analytics be addressed upfront, not as an afterthought or overlay later.
How can AI and analytics enable CSPs to achieve better 5G scale and at the same time reduce their network complexities?
With 5G, there are far more moving parts involved — more connected devices, 5G cell sites, radio elements and antenna, etc….and more data running over the network. We all can get our heads around a 3x increase in network complexity but, with 5G, network complexity will increase exponentially and that demands operators use AI/ML and analytics.
AI-based analytics are essential for automation and providing the in-depth insights on this vast amount of data and numbers of devices, as well as overall network operations. Operators need to be able to not only monitor but predict and address potential issues before they impact their subscribers, and to really understand how each individual subscriber is experiencing their 5G services.
What kind of innovations are being driven by Guavus in this AI and analytics space?
As you know, we’ve been in the market focusing exclusively on CSPs and big data analytics for more than 15 years now, and we’ve had the opportunity to partner with some of the world’s leading operators. They are using our analytics to really improve their customer experience, reduce opex, deliver new types of services to increase revenue and better monetize their big 5G investments.
Last year, we announced our Guavus-IQ products portfolio which builds on our previous solutions and provides operators highly correlated ‘outside-in’ insights on each of their subscribers’ experience and ‘inside-out’ insights on how their own network operations are impacting each subscriber. The products use big data collection, in-memory stream processing, AI-based analytics to ingest, correlate, analyze data in real time.
Within that portfolio, we’ve gotten a number of new operator wins globally with our Service-IQ Device Management Analytics solution. It provides multiple stakeholders within an operator the real-time insights to understand the capabilities and behaviors of the devices on their network. Network planning teams can make data-driven decisions about network rollout strategies based on connected device radio capabilities to plan current investments and future upgrades to their network infrastructure. Product and marketing teams can leverage customer segmentation insights to position the best offers to targeted audiences based on data showing the most adopted device capabilities. And network operations teams can quickly identify suspicious behavior patterns and inform their security and fraud management teams.
We’ve also gotten a lot of interest in our newest member of the family, just announced this month– our 5G-IQ Network Data Analytics Function (NWDAF) product. it provides operators a vendor-agnostic, 3GPP-standard ‘Open NWDAF’ solution that generates real-time operational intelligence in 5G networks to drive closed-loop network automation and service orchestration. As I mentioned earlier, with the complexity and scale of 5G, this is critical.
5G-IQ NWDAF, and our other Guavus-IQ products, provide operators multi-vendor analytics interoperability from their 5G network edge-to-the-core-to-the-cloud. As operators and some of the industry analysts have told us — this is quite unique in the industry – to deliver in-depth analytics insights across all these domains in an operator’s multivendor 5G network environment.
Our SQLstream real-time streaming analytics plays a critical role, as do our partnerships with the leading cloud providers, to give operators the flexibility to deploy our analytics where they’ll achieve the biggest benefits – the 5G edge, core, on-prem or in the cloud. At the Mobile World Congress in Barcelona, we’ve partnered with AWS to demonstrate the new 5G-IQ NWDAF product – it’s the first public demonstration of an NWDAF product and in the cloud, as far as I’m aware.
Where are the big growth opportunities in this space and how do you plan to target these opportunities in the coming months?
We see 5G in all aspects of telecom and adjacent businesses. Private networks and purpose-built network slices will bring advances in traditional telecom services and will enable new collaborations with Massive IoT and the advent of Industry 4.0.
And as I was discussing earlier, a number of operators are looking to deploy private 5G networks for their vertical customers as a new revenue opportunity. To do that, they need to deploy 5G network slicing – AI and analytics are very important to this. By capturing analytics insights from their own mobile network data, combining them with their vertical customer data, they can then package and sell them as new vertical services over private 5G networks to customers in manufacturing, transportation, healthcare, local government, etc.
We’re in a unique position – our AI-driven analytics combined with the verticals expertise and solutions of our parent company, Thales– which spans manufacturing, smart cities, transportation, and other key verticals–gives us the unique opportunity to help operators offer these new value-added services to their vertical customers.