The necessity of these steps is evident when you consider how a slowdown in global data traffic is entirely out of the question – quite the contrary, in fact. The German Federal Statistical Office in Wiesbaden reports that while 40 exabytes (40 million terabytes) of data were transferred in 2014, this figure had risen to almost 140 exabytes per month by 2019 and could soar to over 270 exabytes per month by 2022. The dynamic nature of this development is clearly illustrated by Swedish firm Ericsson: the telecommunications equipment supplier predicts that within the next five years, the traffic generated by mobile data alone could hit 136 exabytes every single month.
The chief driver of this ‘mass movement’ is the Internet of Things (IoT). According to London-based information provider IHS Markit, the global IoT market is set to more than double over the next five years, from around 30 billion devices to a total of 75.4 billion. This is likely to have a major economic impact in the medium term; IoT provider BizIntellia projects that by 2030, IoT could contribute 14.2 trillion US dollars to the global economy.
Of course, all these ‘things’ could theoretically communicate via the cloud. But the sheer volume of data created in this way would very quickly stretch this transport route beyond its limits – in terms of peaks in demand and computing capacity, but also with respect to latency times that would very soon become questionable, if not completely intolerable. Take microtransactions, for instance, where thousands of processes must be executed in a matter of seconds. Even so-called ‘real-time applications’, which are increasingly frequently deployed and deliver the results of data processing within pre-defined time spans, are sensitive when it comes to tolerable delay times, variations in data packet transfer rates and data integrity. This means that they, too, remain fundamentally time-critical applications.
The Internet of Things, Narrow Band IoT and 5G will explode the traffic of the global data highways.
For this very reason, the primary focus of the new 5G global mobile standard is on latency, rather than classic data transmission rates. Achieving less than one millisecond between retrieval and arrival of data is a huge leap, particularly when delay times of ten to 20 milliseconds are commonplace even in today’s best networks. But 5G technologies are broadband by design, making them ideal for transferring large amounts of data when managing industrial plants in real time, for example. 5G will be a true powerhouse, without a doubt – but its power far exceeds what is required for many IoT applications.
This is where narrow-band IoT (NB-IoT) comes into play. As the ‘little brother’ of 5G, NB-IoT radio technology is designed for communications where IoT devices transfer small amounts of data over long periods of time. NB-IoT is not about real-time data transfer, but it needs to work in hard-to-reach places and offer extremely low energy consumption as it transmits tiny data packets across long distances. One of the crucial advantages of NB-IoT will be the ultra-low-power technology that keeps it up and running; such high energy efficiency will extend the life of batteries to up to ten years. Such figures are generally far beyond the reach of technologies such as GSM, 3G and LTE, as neither the networks themselves nor most devices that use them are equipped with the necessary energy-saving mechanisms.
Ultimately, NB-IoT will connect many more ‘things’ to the Internet, while 5G will keep a great deal more data in motion. This means that the technological leap towards the new mobile communications standard will once again trigger a drastic surge in the energy requirements of data centers – which is partly why experts believe that data centers will need to evolve in both procedural and operational terms. Businesses around the world agree on the former: in its traditional role, the data center has already been replaced in many cases by multi-cloud, hybrid and container-based infrastructures. According to a study by Aberdeen and A 10 Networks, 33 percent of companies worldwide have already implemented a hybrid cloud infrastructure, with a further 23 percent planning to do so within the next year. In Germany, Bitkom reports that 32 percent of enterprises with more than 20 employees already rely on multi-cloud computing, rising to 87 percent among large companies with 2,000 or more employees. So how do companies get there?
