If you think 5G will solve all IoT challenges, think again
5G represents the next generation of mobile technology, with GSMA predicting that by 2025, about half of all U.S. mobile connections will occur on 5G networks. With its dramatic bandwidth enhancements, 5G is expected to be 100 times faster than 4G, and is viewed as the seismic shift that will spur adoption of enhanced mobile broadband services as well as IoT, particularly mission-critical use cases. According to estimates, 20 billion devices will be online and connected by 2020.
But even as mobile operators worldwide transition to 5G, the path to IoT will not be obstacle-free. As internet infrastructure providers of all types strive to keep pace with mobile operator investments while protecting their own profitability, we expect some inevitable speed bumps, particularly when it comes to ensuring device performance — low latency, fast connections and superior reliability. Here’s why:
Trying to build Rome in a day: Any time a big project is undertaken in a relatively short timeframe, the likelihood for error and missteps tends to increase. According to McKinsey, mobile operators’ network-related Capex is expected to increase by 60% from 2020 to 2025. Throughput, scalability and reliability will need to increase many times over current levels to support the surge in users, connected devices and traffic volumes brought on by 5G.
Inevitably, this will have a trickle-down effect on external infrastructure service providers, like cloud, CNS and DNS. The entire delivery chain needs to beef up to keep pace with 5G, all while protecting individual providers’ financial interests.
Some providers with more bullish outlooks will dive in quickly, believing their investments will pay huge dividends in just a few short years. Among these companies, we’ll likely see build-outs of size, scope and pace never before seen, to support the anticipated eight-fold increase in global mobile data traffic.
The alternative ‘wait and see’ approach: Other providers will take a more cautious approach, setting certain utilization thresholds that, once met or exceeded, will then warrant additional investment. In the meantime, these providers will reconfigure existing resources, deploying approaches like virtualization to maximize capacity already on hand.
While both approaches have their merits and drawbacks, one thing they share is an increased tendency for performance blips. In the case of the more bullish providers, we’ve seen multiple instances over the years of rapid expansion leading to missteps and errors, some of which seem completely random.
For example, the recent highly publicized Google Border Gateway Protocol misdirect — whereby Google-destined traffic from across the world was inadvertently dispatched to Russia and China — was ultimately the result of an error that occurred when Google expanded its Nigerian presence through a peering relationship with a local ISP.
While the more cautious providers are rightfully focused on preserving profitability, they risk being caught flat-footed as 5G traffic volumes escalate. Additionally, while techniques like virtualization can offer some near-term flexibility, constantly changing partitions and instances can make it much harder to quickly locate growing hotspots.
Comparing a firehose to a cocktail straw: 5G is a last-mile-only mobile connectivity technology. While it will connect massive volumes of users and IoT devices the mobile internet, the speed and throughput advances only apply to this one leg of the packet journey. Even if two communicating devices are connected to mobile broadband via 5G, if the broader internet path between them is restricted or has vulnerability points, the speed or reliability of the messaging will inevitably suffer regardless of the devices’ last-mile connection speeds.
5G will dramatically increase mobile video and media usage, as well as IoT adoption worldwide. We can expect external infrastructure providers to make some inadvertently clumsy steps as they grapple with significant growth demands. This may lead to occasional bouts of added service latency or unreliability. While this may be somewhat tolerable for some consumer-facing applications and appliances, it will not suffice for mission-critical IoT applications, like connected cars, connected cities, medical devices and more.
In summary, we expect to see a few years of awkward adolescence and growing pains as the broader internet infrastructure plays catch-up with 5G. To insulate against performance risks, IoT users and device manufacturers — particularly those of the mission-critical variety — will need to be extra vigilant. The challenge ahead lies in harnessing and analyzing massive volumes of device performance data and proactively identifying growing hotspots across an infrastructure of unprecedented depth and complexity.
All IoT Agenda network contributors are responsible for the content and accuracy of their posts. Opinions are of the writers and do not necessarily convey the thoughts of IoT Agenda.