5G makes network architecture agile, dynamic and cloud native: Infosys

Nitesh Bansal, Senior VP and Global Head, Engineering Services, Infosys, tells us more about how 5G will go about promoting AI, automation and virtualization, and much more. Excerpts:

V&D: How can you transform the underlying architecture in core networks using 5G?

Nitesh Bansal: While 5G can be deployed using the existing core in NSA (Non-Stand Alone) mode, the desired and ultimate target is to have an end-to-end 5G deployment using new 5G core and 5G radio. This deployment, called Stand Alone (SA) mode, will bring the below architectural transformations to the core.

* Service-based architecture enabling agile service creation.
* Functional splits allowing end-to-end slicing.
* Fully cloud native and virtual network functions.

To enable the above, the connected transport and aggregation network also have to undergo transformations. The key changes include bandwidth upgrade to cater 10x speed of 5G, SDN enablement for programmability and Virtualization as much as possible for enabling dynamic services. Further, these need to be orchestrated through proper cross domain orchestration to make it simple for the consuming layers.

Beyond the 5G core transformation, the radio access networks (RAN) is also transforming from small cell-based architecture to micro deployments.

The open network is also fueling this transformation by bringing lot of standards, accelerators and solutions in this area. Service integrators, like Infosys, are very active in this space, and are driving the adoption of these new architecture frameworks without compromising the efficiency or continuity of an extremely large base of existing network services.

5G is a key driver for making the network architecture more agile, dynamic and cloud native.

V&D How will it promote virtualization, AI and automation?

Nitesh Bansal: 5G is pretty much fully cloud native and has virtualized network functions. 5G network require virtualization both at the core and edge. Most of the telcos are in the journey of setting up a common network function virtualized infrastructure (NFVI) to on-board 5G VNFs, and the trend is to make this as a common ‘telco cloud’ across the organization for hosting other VNFs as well.

AI and analytics are built into the 5G architecture to bring higher level of efficiencies in the network. For example, the network is designed to analyse the mobility patterns of the devices and predict the next move of the mobile device so that the network control plane traffic can be optimized.

The architecture also paves way for the network to collect data and expose analytics reports to applications for optimizing the network usage and providing other useful services while hiding the actual data from applications thereby protecting the subscriber privacy.

The agility 5G brings through Service Based Architecture and Slicing un-locks extreme possibilities for AI and ML based automations. For example, trend analysis based dynamic slice adjustments and application prioritization within the slice based on real time data are possible use cases. Even the entire network operations will see a huge transformation by leveraging AI and automation, moving from traditional NOCs (Network Operation Centre) to DOC (Digital Operational Centre).

The key to this transformation is to start with the software architecture thinking, micro service based and containerized to whatever extent possible. Once this is done, we continue to leverage more traditional components where the new technology is still not carrier grade but allow for the possibility to replace as the technology gets ready for prime time action.

V&D: How will 5G empower edge computing and network slicing?

Nitesh Bansal: The 5G standard specifies three fundamental slices namely eMBB (enhanced mobile broad band), uRLLC (ultra reliable and low-latency communication) and mMTC (massive machine-type communication). As the names indicate, these are all based on the specific characteristics that they bring and catering to different use cases.

The key slices that are empowering edge computing are eMBB and uRLLC. With the existing connectivity technologies, many time-sensitive applications like immersive media, gaming, financial applications, autonomous vehicles, etc. cannot be moved to the cloud and edge clouds as the latency impacts the experience.

5G promises 10 times higher bandwidth and 20 times better latency which makes it possible to move these time sensitive applications to the edge (meaning closer to the user) with no experience loss.

5G virtualizes lot of network functions that means there are more general purpose compute in the data centers, including edge data centers. That increases scope for edge computing.

V&D: How can 5G create agile, virtualized environments?

Nitesh Bansal: We have already talked about how 5G is enabling the virtualization infrastructure and agile service creations. I want to bring a different angle to the virtualized and agile environment through the immersive use cases 5G enables.

The high bandwidth and low latency promise of 5G enables immersive and mixed reality use cases possible from anywhere. This means, you can have virtual stores, immersive sports viewing, virtual classrooms, teleporting, virtual and interactive patient care, tele robotics possible without the boundary of location thereby enabling a truly agile environment.

V&D: Which industries are going to move first with 5G?

Nitesh Bansal: There are different answers to this question, but in our analysis below are the key ones:

* Manufacturing – Mission critical & real-time use case enablement
* Gaming and sports – Enabling immersive experience
* Healthcare – Telemedicine use cases
* Transportation – Autonomous and connected vehicles
* Smart city and smart spaces – IoT enablement
* Finance – Edge computing enablement.

High bandwidth and low latency use cases are going to be the first movers. Mobility-related use cases might take little more time as they require a larger ecosystem to be ready. Already telcos are offering fixed wireless and we are already seeing lot of traction in immersive experience space as well both taking advantage of high bandwidth offered by 5G.