Virtualization and the deployment and operation of 5G networks

Telecoms.com periodically invites expert third parties to share their views on the industry’s most pressing issues. In this piece John English, Director of Marketing, Service Providers Solutions at Netscout offers a quick overview of the need for virtualization with 5G.

While 5G undoubtedly holds enormous potential, meeting its demands for increased speed, performance, scalability, and flexible service deployment is likely to result in untenable complexity and OpEx for service operators.

The only truly affordable and practical way to operate a 5G network is to virtualise and automate, change network design, and increasingly manage the network and services from the edge. The application of virtualisation technologies such as NFV and SDN to 5G networks is therefore essential if 5G is to be deployed at a reasonable cost.

Many service providers are already adopting NFV and SDN as a means of boosting efficiencies, launching services faster, and supporting a wider range of applications. Indeed, McKinsey has estimated that the newest technologies in NFV and SDN would let operators lower their capital expenditures by up to 40%, and their network operating expenses by a similar amount.

Are any deployments under way?

Yes. Major service providers in the US, Europe, and South Korea, for example, are well advanced in their testing and initial network deployments. Verizon, AT&T, and Korea Telecom are all utilising the 1Gbps capability of 5G to provide services such as fixed mobile broadband. In addition to enabling valuable enterprise applications, this supports their SD-WAN deployments which, due to their ability to flexibly instantiate new services, are increasingly being seen by service providers as a way of monetising 5G, even at this early stage.

Is that the full extent of it?

Right now, both virtualisation and 5G are being deployed on a crawl, walk, and run basis. We’re still very much in the crawl phase, with most service providers tentatively deploying both technologies in contained parts of their networks and businesses so that they can understand how they work, learn how to manage them and understand how to deploy them most effectively.

With a myriad of new use cases and technologies, fragmented standards around how VNFs are introduced and orchestrated in the network, 5G is filled with unknowns. The relative immaturity of both 5G and virtualisation is therefore currently serving as a barrier to full-scale adoption.

How can we overcome this barrier?

5G and virtualisation each rely on a series of other technologies for successful deployment. Both require tools that enable intelligent visibility into the network in order to generate smart data with clear, actionable insights that will feed automated systems, manage performance, and control automation. Power is nothing without control, after all, and accurate control is fundamental to the management of a virtualised network.

Fortunately, these tools exist in the form of virtualised software designed to gather data, analyse it, and present it in a way that it can be actioned by a service provider’s other systems. Critically, the days of using expensive hardware probes to reactively report on network performance are over. They may still be applicable in trial phases during which a new network is established, but only software-based network assurance will be capable of scaling up to handle the volume of data involved as the scale and scope of a 5G deployment heads towards the running phase.

What will happen when 5G is up and running?

The exact nature of the services of the future is still unclear. Despite all the talk about instantaneously downloading 4K video, or enabling connected cars and supporting a plethora of new devices and services with the Internet of Things, the killer apps for 5G are yet to emerge.

What is apparent, though, is that there are universal requirements that will remain the same regardless of what the future holds. The effective running of services and applications on 5G will rely on visibility into the network, and gaining insights that will enable proactive, as opposed to reactive, responses to any network issues.

 

JGE headshot (002)John is Director of Marketing, Service Providers Solutions at Netscout, focusing on NFV/SDN, IoT/DX, Big Data Analytics and 5G. Before joining Netscout John held product management and marketing positions at Empirix and Tekelec. John has over 25 years of telecom experience covering 4G/3G/2G mobile, cable and fixed line technologies.

Smart data – giving operators the edge

Telecoms.com periodically invites third parties to share their views on the industry’s most pressing issues. In this piece John English, Senior Marketing Manager, Service Providers at Netscout, describes some of the benefits offered by edge computing.

Edge computing is set to play an important part in fulfilling the expectations for 5G and unlocking its promised benefits. The connective tissue of IoT applications and services such as smart vehicles, public safety, remote medicine, robotics, and the networks of sensors deployed across smart cities, will all significantly gain from utilising 5G technology to achieve truly ubiquitous, reliable, scalable and cost-effective device-to-device connectivity.

Moving cloud, compute and processing power to the edge of the network makes it possible to support the ultra-low latency requirements of 5G and IoT applications such as connected cars, which require real time information at all times, while significantly improving the efficiency of devices.

As 5G and the IoT become more pervasive, there will be an increasingly urgent requirement to support the new data loads and unpredictable traffic patterns being introduced, along with the growing demand for reduced latency and increased compute efficiency. It’s little surprise, therefore, that service providers are now beginning to move network infrastructure to the edge.

With IoT solutions underpinning a growing number of aspects in our lives and businesses, it’s crucial that operators have assurance that their connectivity remains ubiquitous, consistent and reliable. But as with any burgeoning technology, there will always be a new set of challenges to face, and dealing with these will require complete visibility across the entire IoT lifecycle.

Intelligence and insight

The benefits offered by edge computing have seen it gain a lot of traction recently, with both established operators and new entrants to the market launching a range of new cloud and server technologies designed to effectively move data center functionality to the edge of the network. As organizations virtualize network components and functions for greater agility, speed and cost-savings, so they will leverage this new NFV architecture to deploy C-RAN (cloud-based RAN) solutions to better support the increase in their customer’s data traffic.

Managed correctly, the data generated by virtual solutions such as these will provide operators with much needed intelligence that will enable them to gain actionable meaningful insights and inform their network policy and traffic management systems. Over time, this flow of information, and the intelligence derived from it, will lead to networks becoming automated and self-optimizing. This will then allow operators to allocate capacity to areas where it’s needed most: whether to manage peaks in network demand or, in the case of IoT traffic, to manage the demands of sophisticated smart city deployments, autonomous cars, or ‘smart’ automated factories.

Use cases such as these are largely unknown territory, however, and the level of data traffic they will generate is unprecedented. Visibility is therefore crucial if operators are to effectively manage their networks. This is particularly the case with the IoT, where operators will be required to report on the status of any mission-critical systems. With IoT technologies underpinning a growing number of critical applications, such as disaster monitoring and military situational awareness, the need for assurance around security and service delivery is paramount.

Though it may be frustrating that a problem with the network may result in a user being unable to access a movie from a Netflix server on their mobile device, the stakes are considerably higher when a similar network problem affects the performance of a remote heart monitor in a hospital.  So, while it makes sense for an operator to reconfigure a network to account for the increasing demands of 5G and IoT services, it must be possible to derive some meaningful insight from the data being produced. If not, it will offer no visibility into what’s happening on that network, and this could lead to very serious consequences.

In simple terms, a frustrated Netflix subscriber unable to watch the latest episode of Stranger Things can contact the customer service center. Automated machines on a production line, however, will not be calling into customer service to let anyone know if they’re experiencing issues with connectivity, which could have a knock-on effect throughout a business’s entire supply chain.

The oil that fuels the networks’ engines

As computing moves closer to the edge, smart data will quickly become the oil that fuels the networks’ engines.

Real-time, scalable meta data imbued with user experience derived from network traffic, smart data offers unlimited scale, across all aspects of the network, both physical and virtual. Once an operator is able to access and analyze this data in real-time, they will be able to gain valuable new insights into how the connected IoT devices and machines on their network behave, how they interact with the network, and the type of traffic patterns they produce. Ultimately, with access to smart data, operators will be in a better position to make more informed decisions about how to optimize their networks, where to allocate capacity, and how to boost performance.

Furthermore, the visibility enabled by smart data can also enable operators to identify anomalies within the network; significant changes that could indicate issues such as network congestion, which could potentially affect hundreds or thousands of connected devices. Actionable intelligence of this type is worth its weight in gold to an operator and its enterprise customers, who rely on the network to provide a communication backbone to support their IoT deployments.

Ensuring that everything is connected

Edge computing may currently be a nascent space, but the benefits of capacity, low-latency and scalability it offers represent great potential for the success of next-generation technologies. As operators prepare for a 5G rollout over the next five years, edge computing will become a critical element of mobile network infrastructure. And with the number of connected devices predicted to reach more than 30 billion by 2020, it will soon be adopted by ISPs, cable companies and a range of other service providers too.

As the industry continues to move network infrastructure to the edge, harnessing NFV and cloud technology to deliver new, faster, more efficient services, the need for visibility across these new-look, complex networks becomes abundantly clear if operators are to make the most of their new investment.

Only by employing a smart data solution will operators have the ability they need to monitor the sheer breadth and depth of the IoT ecosystem, and ensure that everything is assured, and every device connected.

 

Meet NETSCOUT and learn more about their solutions at 5G World 2018, taking place in London, 12 -14 June.

We’re getting closer to the NFV promised land

Talking with Red Hat and Netscout at MWC 2018 reveals that NFV seems to be finally getting there, and not a moment too soon.

From Red Hat EMEA we met Timo Jokiaho, Principal Technologist for Telco (right, above), and Nik Stankau, Business Development Director for Telco (left). Red Hat makes its money from fine-tuning open source software to make it commercial-grade, so it’s in a good position to comment on the state of the software being developed to make everything work – the plumbing if you like – as we move into the 5G era.

They explained that, within Europe at least, France is setting the pace when it comes to NFV development – especially Orange and SFR. Apparently the arrive of Free Mobile a few years ago, and the resulting trashing of margins, motivated the incumbent operators to look harder for efficiencies and one source of those is expected to be NFV.

The flexibility, scalability and ability to innovate that is promised by NFV will be a big part of this. One more specific subset will be Media Function Virtualization, which will offer flexibility when it comes to the use of resources for video provision. This seems to be a refinement of the network slicing concept that will allow dynamic allocation of network resources at times of peak video usage.

As you might expect from Red Hat, they think a big reason for the progress they observe is the non-proprietary approach that is facilitated by open source. Operators are increasingly taking control of the process away from traditional vendors, especially when it comes to orchestration. They also noted that containers are set to be a major factor in the next phase of development of the telecoms plumbing.

Over at Netscout we met Richard Kenedi, President of the New Markets Business Unit, and Petrit Nahi, Chief RAN Scientist. They explained that business assurance specialist Netscout has been accelerating in the telco direction with the help of acquisitions such as Arbor Networks. Thanks to its network and service assurance work Netscout feels in a good position to comment on the state of the network.

They confirmed that NFV is definitely getting there and that the increased focus we’re seeing as 5G gets nearer is forcing practical action upon the industry. Initially CSPs are likely to focus a lot on capacity, and fixed-wireless access seems to be getting a lot more traction thanks to good R&D on millimetre-wave propagation. Meanwhile the low-latency part of the 5G equation is leading to a lot of talk about mobile-edge computing.

The third pillar is IoT and Netscout confirmed our previous finding that it’s all about NB-IoT now. They confirmed that there is a lot of CSP activity around NB-IoT and that the activity is largely moving from B2C to B2B implementations and business models. They also said network slicing will be a critical factor in taking IoT to the next level. Lastly they confirmed the vibe we got from a lot of other people at the show that the overall market will bottom-out this year and then start to pick up.

As with so much else at MWC 2018 the perspective from a couple of companies involved in the plumbing of the network was a sense of ramping and practical measures being taken in preparation for 5G in 2019 and beyond. Which is just as well as, unless we get that stuff right, all the clever New Radio in the world won’t amount to much.