Has the internet reached its full potential?

Mobile & Apps

Predictions assume that in the future half of the global volume of data will no longer be generated by or between people, but rather by vehicles, sensors and various types of networked devices. And the data volume continues to increase dramatically. Telecommunications equipment supplier Ericsson estimates that global data traffic will increase five-fold over the next five years to 136 exabytes per month compared to today. The expectation is that a typical end user generates up to 1.5 gigabytes of data per day, a vehicle with around 4 terabytes already several times that.

The 3GPP body responsible for standardization in mobile communications already had this development in mind when it defined the fifth generation of the mobile communications network. 5G was essentially about offering significantly higher data rates (5G should deliver 100 times the throughput of LTE), enabling latency times in the millisecond range and offering significantly higher availability and reliability of the network.

5G use cases and the role of edge computing

The billions in investments required to build 5G networks represent a big bet for the telcos in the future. This applies to both services for end consumers and the support of new IoT business models (Internet of Things). A distinction is typically made between four main use cases for 5G technology:

  1. Super broadband experience for the mobile user (with 8K video streaming, high definition mobile gaming and virtual reality applications);

  2. Wireless stationary broadband connections with a throughput of well over 100 Mbit / s;

  3. IoT connections, for example for household appliances, building control or the intelligent cities of the future;

  4. Time-critical control systems, for example in autonomous driving and in the automation of industrial plants;

However, 5G technology is only one of the important prerequisites for the implementation of such smart services and products. The significant increase in the number of end devices, the explosion in the amount of data exchanged and the need for minimal latency times place new demands on the transport networks of telcos, but also on those of companies that 5G technology alone cannot meet.

The number of so-called edge devices, such as mobile end devices, vehicles, household appliances, drones and infrastructure components, should increase dramatically in the medium term: According to estimates, 75 billion of these devices will be in use worldwide in five years. On the one hand, this creates new challenges in coping with the amount of data that arise, on the other hand, today's end-to-end runtimes (roundtrip time) of more than 50 milliseconds between the mobile device and the central cloud (or central data center) limit the implementation of time-critical applications, such as in industrial robotics control.

The edge computing concept provides a solution here. The idea behind it is simple: You bring the cloud with its computing capacity closer to the user and thus achieve runtimes of less than 10 milliseconds - but without depriving it of its advantages (scalability, agility, payment by usage, distributed resources, innovative strength).

Data preprocessing between the end device and the cloud

Edge computing also means that it can make sense not to locate these high-performance computing activities in the end device itself (vehicle, machine, infrastructure component, drone). An edge computing device is so close to the place where the data is generated within a network that response times of a few milliseconds can be achieved.

It acts as a (pre-) processing hub between the end device and the cloud and provides a secure and standardized environment for the execution of customer-specific logic. In this device, data is processed and / or compressed according to a predefined business logic and the consumer receives direct feedback if necessary. This ensures that mobile end devices do not become more complex at the same time, despite the ever increasing scope of functions, and that energy consumption, weight and, last but not least, costs can be minimized.

The processing of vehicle data in the context of autonomous driving is often mentioned as a relevant application of edge computing. Cars stand idle for an average of 96 percent of their lifespan - this is where the load-sharing advantage of edge computing becomes very clear: You don't necessarily have to install expensive components in every vehicle, but shift the logic to an edge computing device that can serve a large number of vehicles.

The potential for weight and cost reduction through edge computing devices can be illustrated, for example, by the complex image processing and video analysis of camera drones. And the application scenarios can be expanded almost infinitely: from the implementation of blockchain logic in production, logistics and usage chains to the complex calculation of augmented / virtual reality models, for example for system technicians during installation or maintenance work, to real-time control of production systems .

It should be noted here that 5G technology does not necessarily have to be used to bridge the “last mile” to the mobile device. Alternative technologies such as the new WiFi 6 standard, which has similar performance characteristics to 5G, are also conceivable in industrial plants.

Outlook and success factors

The 5G expansion is being strongly advanced globally and the numbers are impressive: Goldman Sachs expects investments in the three-digit billions (dollars) by 2025, of which 150 billion dollars will go to China alone. These investments flow first into the acquisition of spectrum licenses and then into the expansion of radio networks and mobile cells, fiber optic networks and an agile, software-defined core network. The telecommunications companies will constantly adjust their investments and the replacement of their 3G / 4G components in relation to market acceptance, the competitive situation and the additional revenues that can be achieved with 5G services.

Edge computing will also grow strongly in the slipstream of the 5G expansion. Gartner's market researchers expect that in future 75 percent of data will be processed “on the edge”. As a result, annual growth rates of 50 percent are predicted for the edge computing market. In 2025, Grand View Research expects a market volume of almost 30 billion dollars. So it's no wonder that cell phone providers are also working on appropriate edge computing products and services.

The success of the carrier's efforts to position itself as a central provider in edge computing will largely depend on the extent to which standards are established and accepted that allow simple deployment of company-specific applications on the mobile edge. At this point, open (open source) standards must be used so that edge workloads can be installed in a "carrier-agnostic" and internationally standardized manner. One example of this is the ONAP standard promoted by AT&T and China Mobile. And Deutsche Telekom is also positioning itself with the launch of a corresponding open source initiative.

In the field of edge computing, in addition to telcos, there are also cloud providers with their sometimes competing approaches, such as Microsoft AzureIoT Edge, Amazon's AWS Outpost and Greengrass or Google's Coral and Edge TPU. There is potential here for possible cooperation models between the carriers (strength: presence in the area) and the public cloud providers (strength: ecosystem, financial strength, innovation and software development) - corresponding options for this are already being evaluated at various points.

This is what companies and users should do now

The interaction between 5G and Edge Computing opens up new possibilities and areas of application for companies. 5G as high-performance mobile communication technology and edge computing with its possibilities for the development of new smart services that are independent of the network technology used (e.g. 5G, WiFi 6, NB IoT, LoRaWAN). So it is high time to deal with the topic, the currently available technical concepts and the provider landscape.

Companies should also look into the analysis of potential business models and value propositions that 5G and edge computing enable. Within the IT and digitization departments, it is important to deal with the further development of the enterprise architecture around edge components and to gain initial experience in the planning and implementation of corresponding pilots as soon as possible. (mb)