In 2017, roughly 108,000 people came to Barcelona to attend Mobile World Congress (up 7% from 2016), where the overall messaging was mainly focused on just two topics: 5G and the IoT. To put this in perspective, in 2016, the mobile industry started to talk about the convergence of the IoT, cloud and 5G; the year before was all about SDN and NFV.
This year, 5G took center stage and, for the first time that I can remember, car manufacturers showed up at MWC to demonstrate their connected and autonomous cars that 5G should “easily” enable.
All of the major mobile operators focused on providing IoT solutions for verticals like automotive, public transport and logistics, health and wellness, smart city and utilities, agriculture, media and entertainment, and manufacturing. Specifically regarding manufacturing, the focus was on the so-called “fourth industrial revolution,” or “smart factory,” that 5G should enable, likening the increase in bandwidth to prior industrial revolutions triggered by the steam engine, electricity and automation. Another very promising use case for 5G is as a replacement for current, and expensive, “last-mile” technologies based on fiber.
The landscape was littered with demos of virtual reality headsets for gaming, autonomous cars, and small robots (playing drums and also performing other slightly more useful activities) connected to a “central intelligence,” but I came away with the feeling that there is not much substance behind these use cases. The mobile industry is pushing the need for 5G (so it can sell more equipment), but the business models that should support such an expensive transition are still unclear. Most agree that 5G will require NFV and SDN technologies in order to enable the creation of a flexible network capable of providing the required capacity in the right place and at the right time: a fully dynamic, elastic environment.
The benefits of 5G, beyond the pure increase in throughput, are clear in terms of guaranteed latency, device density (billions of IoT devices), coverage and elasticity—but without a well-defined “ecosystem” of network equipment, applications, services and mobile devices, the investment risk may be too much for most operators. Also, remember that “4.5G” and/or “4.9G” technologies like LTE-Advanced and LTE-Advanced Pro have not been yet been fully exploited, and are already available from all the major network equipment manufacturers.
Another roadblock will be the standardization process: 3GPP will not finalize 5G specifications before 2020 but, despite this, South Korea (KT, SK Telecom and LG U+), Japan (NTT DoCoMo) and China (China Mobile) are driving toward 5G because they need to replace narrowband Internet and expand broadband service coverage now. In the U.S., AT&T and Verizon have also announced aggressive plans. Currently, 24 operators have tested 5G in laboratory settings, 27 have announced the intention to start lab trials, and 12 have reported conducting field tests. The risk is that not all pre-5G technologies will fit the future standards and, honestly, the 4G technology (LTE) available today should allow the implementation of many IoT applications.
With all that said, I believe 5G will definitely come together with full-scale IoT, but the timing and availability of real applications still need to be defined.
We work with a range of mobile service providers delivering solutions tailored to meet their needs. Empirix has also written a whitepaper that explores the critical importance of true modelling and runtime capabilities to achieve adequate customer experience testing.