IoT & Virtualization Go Together Like a Horse & Carriage

IoT & Virtualization Go Together Like a Horse & Carriage

At the 2017 Mobile World Congress (MWC) in Barcelona, I noticed several marked differences from the previous two years that I have been covering IoT for Heavy Reading. One was that nearly every company had some form of IoT demonstration on its stand. IoT also seems to be central to every company’s strategy going forward, irrespective of whether they are a chipset, device, radio access network (RAN), core network, platform or BSS/OSS provider. Additionally, some of the lead IoT use cases were at the show engaging with vendors and operators — primarily connected car OEMs, but also smart city applications. However, what was lacking from all of these interactions was a clear monetization strategy across the entire mobile value chain.

The GSMA and the NB-IoT Forum kicked off the conference events. Their key message was that, after a whirlwind standardization and testing process through 2015 and 2016 that broke all known speed records, 2017 was about deployments and implementation. With 60 operators supporting the NB-IoT initiative, together with major chipset manufacturers and module companies, there are 25 commercial launches planned in 20 countries for 2017, providing the basis for both momentum and scale. To support this ecosystem, the GSMA created the Mobile IoT Innovators community that currently has 161 members worldwide: 35% in Asia-Pacific, 35% in Europe, 25% in the US, and 5% in the rest of the world.

The low-power, wide-area (LPWA) radio access space is similar to the early days of cellular, with multiple competing standards scrambling to win favor with operators. The GSMA M-IoT initiative supports NB-IoT, LTE Cat M1 as well as EC-GSM, which can be deployed on GSM networks. These standards all improve coverage and reduce battery consumption, while making more efficient usage of the spectrum and networks to deliver varying degrees of bandwidth for differing use cases. Key proponents of NB-IoT are Vodafone, Deutsche Telecom (DT) and China Mobile, while AT&T and Verizon are driving LTE Cat M1 in the US, which has bandwidth advantages for handling some video surveillance traffic and other use cases. Both US operators claim to be introducing NB-IoT in 2018. The most vocal proponent of EC-GSM is Orange, which has significant GSM network properties in Africa. In fact, the most recent GSMA mobile economy report shows that 46% of current global subscribers are still GSM, and that by 2020 GSM will still be in existence, at around 28%.

Making this LPWA space even more complex is the continued growth of alternative technology solutions for machine-to-machine (M2M) and IoT applications. The LoRa Alliance, SigFox and Ingenu, together with Telensa and Silver Spring Networks, have all made progress in the last year. These proprietary or semi open technologies all take advantage of unlicensed spectrum. And so, it was not surprising that, as part of the development work for the planned LTE Release 14, Ericsson and Huawei proposed extending NB-IoT into the unlicensed spectrum.

For M2M customers in developed markets, the landscape has gone from bleak in 2014, when the prospect of losing GPRS-enabled M2M devices became a nightmare scenario, to a brighter world full of competing technologies that are, in fact, continuing to multiply. In the US, the recent release of 3.5GHz Citizen Broadband Radio service (CBRS) and the development of private cellular LTE solutions could jointly redefine the IoT business model. Companies, such as Quotus in the UK, believe that private cellular is a service model that will be enabled through the use of shared spectrum. This could lead to the prospect of private companies not being reliant on traditional operators for access to the LTE-based LPWA, simply because they are gatekeepers to spectrum.

This then focuses attention on the monetization aspect of the equation. The fact that customers now have access to a range of alternative technologies means that cellular operators must be more responsive to their customers’ needs, and tailor services to their specific IoT requirements in order to win their business. They must provision and manage devices in an agile manner, flexibly adjust services to changing customer operational requirements, and accurately cost out and bill these LPWA microservices in a way that is competitive as well as attractive to customers. For these reasons, I walked away from MWC with the strong belief that IoT is the customer use case that virtualization has been waiting for.

This belief was reinforced when I read the Mobile World Live annual survey, published at MWC, that highlighted the lack of traction achieved by virtualization. Only 45% of respondents said they think virtualization will bring costs down to a sustainable level, and 43% were not sure this will happen. I believe that this lack of consensus on the need for virtualization to help operators redefine their costs base can be overcome with a shift of emphasis from cost only to cost and revenue focus. The only way that operators are likely to make money with IoT is by controlling costs, and pricing microservices to meet emerging use cases. Virtualization is essential for both.

The explosion in LPWA technologies, and the emphasis that operators are placing on their deployment, should be the boost that virtualization needs. Additionally, since NB-IoT is the first occurrence of network slicing on a RAN, the lessons that operators learn from deployment should prove invaluable to understanding the practicalities of broader network slicing in 5G. However, this knowledge and success in IoT can only be achieved if operators embrace virtualization. That is why IoT and virtualization is truly a horse and carriage type scenario — one is dependent on the other. (See IoT & 5G Are Hot Trends, But What Is the Business Opportunity Now?)

Article originally published on Light Reading.