Insights: What will be the game changer app for 5G networks?
Following this year's Brooklyn 5G Summit, I'm still wondering whether there will be a "killer app" for 5G networks. For 4G LTE, it's easy to argue that the most important use case turned out to be video, which went from about zero on earlier generations of wireless networks to making up the majority of traffic.
But as in past years, I again came away convinced that 5G will be much more important for industrial and IoT use cases – and better network management for the providers – rather than leading to a huge breakthrough for consumers. Still, simply providing enough capacity for smartphone users does seem to be a worthy goal. At the conference, a number of presenters talked about the use cases they envision for the technology.
Jongsik Lee from KT talked about how 5G was used at the recent Winter Olympics in Pyeongchang. He focused on five different trials, which all used pre-standard technology.
Lee said they were able to demonstrate up to 20Gbps connections with less than 1ms of latency using the 28GHz band, with network equipment from Nokia, Ericsson, and Samsung and devices from Intel and Samsung.
Specific tests showed a 5G tablet receiving content at 3.2Gbps; an "Omni-View" app that tracked a cross country skier; "sync view," which provided a first-person view broadcast over wireless (inside the track; bobsleigh, short track, etc.); interactive time slices, which stitched together massive camera views; a connected car demo; and a drone torch relay.
In Lee's view, the tests showed that mmWave technology can increase 5G coverage in a cost-effective way, and that 5G repeaters can enhance outdoor-to-indoor coverage, though they require more investment. Lee said he expected three carriers in the country will offer 5G service, and that the first "killer device" will be a mobile smartphone with unlimited mobile data. Later, he expects to see mobile VR, new immersive media, and enterprise private networks.
"5G may be disruptive" to the manufacturing industry, according to Bosch's Andreas Mueller. He said the combination of edge computing and network slicing will be key technologies for the factories of the future.
Mueller talked about the many advantages 5G could provide – including allowing much more flexible placement of equipment with ubiquitous wireless connectivity – but stressed that many of the industrial requirements haven't been fully addressed yet and may need to wait for further iterations of the standard. But if it all comes together, it's possible that "Industry 4.0 may become the killer application for 5G," Mueller said.
In another talk, Dina Katabi, an MIT Professor, discussed using 5G for health care, VR, and smart cities. For health care, she talked about the problem of seniors falling, and said that while we have pendants that seniors can use to notify people if they fall, seniors often won't wear them. Instead, she talked about ways of monitoring falling using radio signals alone.
For smart cities, wireless systems could detect the position and speed of cars very accurately, and locate parking. In VR, Katabi talked about streaming 6.5Gbps of data to a headset, and said that existing Wi-Fi and cellular systems aren't capable of this, but that mmWave systems with self-configuring mirrors might be.
How 5G Networks Need to Evolve
At most technology conferences, I find myself to be one of the few people who point out that, despite all of the focus on technology, over the past decade, actual productivity rates have fallen; I believe this is the biggest issue the tech industry needs to address this year. So I was pleasantly surprised to hear Marcus Weldon, President of Bell Labs and CTO of Nokia, address the productivity issue, and explain how he thought 5G could contribute.
Using statistics from the Technology CEO Council, Weldon pointed out that while digital industries have seen average productivity growth of 2.7 percent over the past 15 years, the physical industries have seen only 0.7 percent average productivity growth (in the same period).
To fix this, Weldon focused on the efforts made in a much broader range of jobs – including the physical industries – to collect data, process data, and perform "predictable tasks." He said that new technologies, including 5G, offer the potential for much more automation.
To make this happen, Weldon argued, we'll need a "new value architecture" that starts with a smart network fabric built on top of devices and sensors as well as the cloud. On top of this will be a variety of services, including the core network, a programmable network OS, AI services, and finally, "digital value platforms" – the actual applications. And, he continued, we need new means of ensuring data security to make it all work.
Marc Rouanne, President of Mobile Networks for Nokia, said that for the next industrial revolution we need a radical new vision – rather than classical improvements – to "change the way humankind interacts with machines and knowledge."
Rouanne made the case that 5G is coming even faster than predicted, and said that enhanced mobile broadband services would begin to roll out next year, followed by applications designed for "the fourth industrial revolution," with more of an emphasis on digitization, automation, and more flexible production.
Driving these new applications will be "data democracy" and a new emphasis on openness, including open networks, orchestrations, and AI engines within the network. One of the big changes Rouanne envisions involves the way 5G is requiring operators to rethink their network architecture to make it faster to bring new services to market. To that end, he pushed Nokia's ReefShark chipset as a technology helping to enable breakthrough network performance and cost reduction.
Rouanne talked about technologies like mmWave networks with beam-forming and massive MIMO antenna systems, as well as enabling lots of different spectrum variants, each with its own properties. He focused on network slicing, and said that instead of running a private network for a particular application, you could create thousands or millions of "dynamic slices" in a network – and do this in milliseconds.
Applications he discussed included a 5G smart seaport, which would involve using sensors to manage traffic lights; real-time video and virtual reality; vehicle to anything ("V2X") applications to enable automated driving; infotainment, maps, and public safety using different slices of the network; and industrial applications that would use "ultra-reliable, low-latency connectivity" features for things like controlling interactive mobile robots.
Rouanne also hinted at future evolutions of the 5G standard, and papers presented at the summit featured ideas such as using 90 GHz radio spectrum; "edge clouds" to enable remote haptic experiences; and turning light fixtures into a 5G network for in-building services.
I'm not sure whether individual mobile phone users will notice any huge jump when 5G rolls out, but it does seem that there could be many interesting applications for the technology, and it will certainly help keep the networks ready for the massive volume of data that we're already using.