Nokia and AWS partner to rule the edge cloud, but where do MNOs fit?

A new and highly strategic alliance between Nokia and AWS highlights the new power politics of a world in which telecoms and IT are colliding and new revenues will be heavily driven by the distributed cloud and the edge network. The companies will cooperate in many areas from SD-WAN services to edge cloud reference architectures to IoT applications – in the kind of partnership which puts the MNO’s hopes of a brand new 5G business case at risk.

It was inevitable that the webscale giants would want a key place in the telco network as the cloud moved to be fully distributed to every user and even connected object. But extending their control from their data centers right to the edge was not a guaranteed victory. Their ambitions were challenged by the mistrust of the public cloud to support business critical services; and by operators’ own dreams of becoming significant cloud services platforms. As enterprises increasingly moved flexibly between different clouds, there was the risk, for AWS and Microsoft, that they would become commoditized, while the real enterprise control passed to the network linking the clouds to the edge resources and the users.

Amazon’s AWS has recently moved decisively to counter the risk of the commodity role by striking strategic partnerships with major players in the mobile and telecoms space – AT&T was a key alliance, and now Nokia. In the case of AT&T, the operator seemed to recognize that, while its mobile and fixed enterprise networks – and its evolving on-demand platforms – are trump cards, it still cannot hope to rival AWS by building its own cloud. In the case of Nokia, this is another way in which the vendor may be able to sideline its core MNO customers and seize the distributed, on-demand enterprise business for itself.

The agreement is not exclusive but it is highly strategic. Initially it is billed as a means of making public or third party clouds more attractive to enterprises and operators than their own private platforms, making the AWS and Nokia clouds easier to use whether to launch and deliver services, or for internal back office functions.

But the deal goes well beyond that and fits in with Nokia’s current preoccupation with edge computing, an emerging trend in which the Finnish firm aims to take the lead, as part of its mission to increase its revenues from vertical markets and the Internet of Things. With AWS, Nokia will be better placed to enable service providers in edge computing, and also to offer a seamless SD-WAN based on a combination of the Nuage platform which Nokia acquired with Alcatel-Lucent, and the AWS clouds, according to strategic cloud investments leader Prabir Datta.

The announcement is short on details but would see Nokia providing support services for an operator or connected enterprise to virtualize their applications and internal functions, and move them to the AWS cloud. Those services will include “consulting, design, integration, migration and operation for infrastructure and applications”.

The four key objectives are:

• Nokia will support service providers in their AWS implementation strategy with consulting, design, integration, migration and operations services.
• Nokia and AWS will work together to generate new 5G and edge cloud strategies and guidance for customers including reference architectures.
Nokia and AWS are working to improve the user experience for Nuage Networks SD-WAN customers who use AWS, allowing them to manage both platforms from a single pane of glass.
The partners will work to commercialize IoT use cases based around combinations of AWS Greengrass, Amazon Machine Learning, Nokia Multi-access Edge Computing (MEC) and Nokia’s IMPACT platform.

Datta said: “The goal is to simplify the headache of the service providers in the sense that they don’t have to invest in private data centers going forward. With AWS it will be easy for them to think of applications which could be easily migrated, along with network operations to AWS cloud.”

As so often when it comes to the cloud, Nokia is initially positioning itself as the friend to its telco customers, but it is not much of a leap to see the company, helped by allies like AWS, going directly to other types of enterprises – some of which operators are also hoping to target with distributed cloud migration services. While companies like AWS may lack the critical link in the chain which is connectivity, Nokia does not have that weakness. Its Wings platform for IoT has already shown how it can provide connectivity from core to edge, which is provided over many operators’ networks, reducing them to the dreaded bit-pipe.

Network slicing will enhance the model further, allowing a provider to activate virtual slices of capacity, on-demand, for individual enterprises or MVNOs (or eventually consumers). The power in that situation clearly rests with the company that can control the slice engine – the cloud-based mechanism which calls up the slice; orchestrates all the virtual and physical resources it needs, including airwaves, air interface and transport; and provisions it with its own policy and core network functions, to meet the needs of the particular service (whether that is low latency, high availability, high data rates, or all of the above).

There are three potential power players in the slice engine – the operator, harnessing its control of the network; the cloud provider; or a third party which manages to work with both these stakeholders but drives overall management of the slices. Such an organization would implement the slice engine (probably based on open source technology but with heavy degrees of optimization to offer the maximum flexibility and dynamic response). And it would control the commercial agreements with the slice customers, and be able to layer professional services, such as enhanced security management, around that.

We can clearly envisage a company like Nokia targeting this high value spot in the on-demand chain, building on its established relationships with network operators and its newer bonds with cloud providers, especially AWS. It can present itself as a multivendor platform, since slicing – especially for truly dynamic services and multinational reach – will need to tap into multiple networks and probably multiple clouds. Telcos and cloud providers could go multivendor too, engaging in ‘co-opetition’ and roaming deals with their rivals, but a third party could avoid some of the complexities of that scenario.

So there is all kinds of significance to Datta’s apparently throwaway comment that “we are currently working with multiple operators to actually test out the spaces we have been talking about.”

This dynamic, multi-operator, sliced platform is a long term vision, but Nokia does seem to be putting the kind of foundations in place which could give it a headstart in network-on-demand and eventually slicing services. Even if that comes at the expense of sidelining its traditional MNO customers, it will help to achieve the core goal of increasing its addressable market, especially in the IoT and in the verticals with the greatest need for specialized high value networks (Nokia has already announced contracts with mining, utility and public safety organizations, all of which fall into this category, whether supporting humans or ‘things’).

In the nearer term, Nokia – and other companies, like HPE, which may feel able to sit in the high value spot between the cloud and the network providers – can also take advantage of the fact that most operators are moving away from running their own private clouds and are starting to work with AWS or Microsoft.

In some cases, the operator will have the clout to work directly with a cloud partner and squeeze out a would-be orchestrator like Nokia. For instance, AT&T has a clear chance in the on-demand enterprise world via its year-old alliance with AWS. This showed organizations recognizing that they cannot support the entire distributed cloud chain single-handedly but can combine their skills to provide enterprises with a rich set of solutions. Radio specialists need IT partners – and vice versa – to succeed in the converged market, and so AT&T and Amazon will integrate their respective networking and cloud capabilities. This will start by expanding existing services – connecting devices to the cloud, and working on optimizing those links; but also preconfiguring sensors and devices for efficiency in the IoT, and working on overall platform security and threat management.

This showed AT&T acknowledging a reality which most carriers will have to do too – that they are not in a position to compete with Amazon AWS or IBM directly in offering cloud services. But they have highly valuable expertise in device connectivity and management, and in provisioning and monetizing large numbers of gadgets and consumers.

So alliances like this one are sure to proliferate, though some telcos will be more successful than others in avoiding a bitpipe role in the cloud, and securing a significant role in the value chain when they join forces with Amazon, Microsoft or vertical market platforms like GE’s Industrial Internet Initiative (in which AT&T is also the primary carrier member). Some, however, will lose the advantage of connectivity and network services expertise to Nokia or other telco vendors, or may choose to take a smaller, but less expensive, share of the value chain by partnering with a Nokia/AWS combination.

AWS and Nokia are particularly focused on the edge computing space, which Rethink’s latest operator surveys show is of high interest to mobile operators, but is also seen as very high risk to deploy, because the architectures and revenue models remain uncertain. Relying on an infrastructure and managed services partner would significantly reduce that risk while allowing the MNO to offer edge-oriented services to its customers.

The ambition of the Nokia/AWS agreement is especially clear in this area. The companies plan to work together on a reference architecture to provide both service providers and enterprises with a blueprint to implement an edge cloud strategy. There will be inevitable tensions as they work this out, since both companies have their own edge clouds built around different assumptions.

Amazon has been investing in its Greengrass technology to drive a full developer and services ecosystem around the edge, particularly for the IoT. Nokia was an instigator of ETSI’s MEC (Multi-access Edge Computing) working group and pushed the idea that the edge cloud resources, such as processors and storage, should be collocated or fully integrated with base stations – as small cells proliferated to support densification, so those resources could be pushed further to the edge, expanding from central offices to macro sites to small cell sites.

In fact, Nokia’s thinking seems to be out of vogue with many operators now and there is a new focus on architectures, like that of the OpenFog Consortium, which place compute resources anywhere in the network, from center to edge (which may even be a device or home), depending on use case. It is possible that working with AWS will help Nokia adjust its own edge cloud platform to support the option of a less telco-centric, more IT-driven view of the world.

“We are in the process of building and thinking about this. We do not have a clear idea yet,” Datta told LightReading. “We don’t know whether it would be an AWS-based edge cloud solution or it could be a mix of both because Nokia has its own edge cloud. It could be a combination of both at some point but at this point we are still building out the reference architecture and it is at an early stage.”

In addition, Nokia should reap immediate benefits from the SD-WAN collaboration, allowing customers of its Nuage Networks SD-WAN to manage those services alongside their AWS cloud services, in a fully integrated way.

And there is an IoT element to the deal, of course. The companies will cooperate to build a series of pre-packaged IoT applications which will harness both Amazon and Nokia capabilities to reduce risk and time to market for customers such as cities, telcos and vertical industries.

The technologies involved are strategic to both – Greengrass, Amazon Machine Learning, Nokia’s implementation of MEC and Nokia’s IMPACT platform. When Amazon announced full availability of Greengrass in June, it was clear it could be a lifeline for MEC, which was starting to be eclipsed by OpenFog and other IT initiatives (see Wireless Watch June 14 2017 ‘Amazon Greengrass may provide lifely for MEC and the cellular IoT’

We posited that Nokia would do well to embrace the Amazon edge/IoT technology, and that is now happening. AWS Greengrass provides edge computing capabilities for IoT applications over fixed or mobile connections. Given the market weight of AWS, this could accelerate commercial projects more effectively than the collegiate efforts of ETSI’s MEC initiative, or the OpenFog Alliance, on their own. But these industry cooperations could ride on Greengrass to help achieve their own objectives – especially as Microsoft, despite significant effort behind Azure IoT and in cloudlets, has yet to make its edge component generally available.

However, Microsoft has recently enlarged its alliance with Cisco, the leader of OpenFog, allowing customers to build and host IoT applications in Azure, and then distribute them to the edge via Cisco’s Fog technology. So it is perhaps natural that the most MEC-focused of the major vendors, Nokia, should seek out the other major cloud platform and try to tie AWS efforts more closely to MEC.

Amazon AWS hopes to insert itself into the distributed content and IoT value chains at every point with Greengrass. The CTO of AWS, Werner Vogels, said in July: “With AWS Greengrass, we can begin to extend AWS into customer systems—from small devices to racks of servers—in a way that makes it easy to do the things locally that are best done locally, and to amplify those workloads with the cloud.”

Amazon described how “Greengrass brings cloud programming and functionality to sets of IoT devices – empowering them to communicate and react when a cloud connection is not possible. These devices together are known as a Greengrass group. Groups are defined and configured from the cloud.”

Every group needs a Greengrass core to be created in order to function, and this software is installed on a physical device, such as an edge server or small cell, and connects it securely to the AWS cloud. The group can then be extended and defined in the cloud before being deployed. The idea is to make it relatively straightforward to distribute the cloud for certain applications, and that is where the hopes for MEC and fog lie. If the base of AWS developers is encouraged to build edge compute applications, and they find it easy to deploy the Greengrass cores on MEC or fog appliances, the spread of either approach could be greatly accelerated.

Nokia’s Thorsten Robrecht said: “For years I’ve observed the convergence of IT and telco and believe that cloud services are one of the areas where the two paths have definitely crossed. The combination of Nokia MEC and AWS GG distributed cloud capabilities will bring significant advantages to the Industrial IoT ecosystem – to IoT sensors, devices and application developers.” The MEC/Greengrass combination has already been in tests in Nokia Labs, targeting use cases like mining or oil rigs, combining edge compute with a localized, self-contained RAN of small cells managed by a virtualized evolved packet core, on-premise or in the cloud.

Robrecht wrote: “KPI performance figures from Nokia MEC and AWS GG testing in Nokia Lab are very promising: 93% of messages were processed at the edge of the network, round trips time decreased by 28% and latency was 39% lower compared to the case where data needs to reach the centralized core.”

Of course, Greengrass will be deployed in many environments which do not require optimized cellular connectivity or core, but also in many which do, which is where the hopes for MEC come in – and potentially for the mobile operators which control those cellular networks. But the more slicing turns the edge cloud into a multi-RAT affair, and the more cellular technologies are deployed in unlicensed or shared spectrum via innovations like MulteFire, the more precarious the MNO’s position becomes, unless – like AT&T or SKT – it invests very heavily in IT, cloud, edge and services capabilities, and has the market weight to attract strategic partnerships with the cloud giants.

Otherwise, there is a gap into which Nokia and other vendors will try to push. While Amazon is working closely with some MNO partners on mobile cloud developments, it is clear that, in a scenario where shared spectrum and private networks have reduced the power of the mobile operator, AWS would be a logical company to step into the MNO’s shoes, aided and abetted, perhaps, by the MNOs’ former friends.