6G Mobile Communications

Everywhere, anytime, and ubiquitous – for mobile communications to fulfill the promise of universal availability, it has become clear over the past few years that a restriction to terrestrial connectivity is not sufficient. The “Non-Terrestrial Networks“ of the fifth generation of mobile communications, which integrate satellites into the communications infrastructure, will be further diversified and expanded by an additional dimension with 6G. Ground, air, and space combine to form multi-layered 3D networks, for instance, incorporating drones, airplanes, or High Altitude Platform Stations (HAPS) into mobile communications in addition to satellites. This increases its resilience to external influences, but with every additional layer and element, the need for coordination also grows.

We are addressing this challenge in various ways. To replicate the processes and structures of the highly dynamic and complex 3D networks, we have developed a 6G-capable system-level simulator that can evaluate potentials and limits in practice-relevant scenarios. Furthermore, we investigate how terrestrial and non-terrestrial mobile communication systems can be integrated to coexist within the same frequency bands and enable seamless connectivity even in remote areas. Here, we develop AI methods and simulation tools that can harmonize the interplay of multiple radio systems while taking the open O-RAN architecture into account. In addition, we are researching new waveforms for 3D networks that are so energy-efficient that even capacity-limited satellites can process high data rates. 

Combating climate change is one of the central tasks for humanity in the 21st century. The Paris Agreement and the European Green Deal commit the economy and society to become CO₂-neutral by 2050 at the latest – an ambitious goal to which mobile communications must also contribute. Although wireless communication becomes more efficient with each generation, 6G risks – as is the case with 5G – losing these benefits as the demand for data and thus the need for electricity will grow rapidly in the 2030s. Consequently, accompanying measures must be developed that decouple data volume from energy consumption. The formula is: More performance, less power.

Decarbonizing mobile communications is one of our core concerns. We apply targeted methods where particularly significant effects are expected. Therefore, our research activities primarily focus on base stations, which are responsible for about 80 percent of the energy consumption of networks. The key lies in energy savings: The base stations are put into different states of wakefulness and sleep depending on the need, so that the activity of mobile communications and the associated energy consumption can be throttled at any time. Since there are various interdependencies between base stations and devices, we are also working on optimally coordinating their sleep modes. This ensures that the impact on users remains minimal. To identify, assess, and refine the most effective methods in this area, we are working on a 6G-capable simulator for energy savings. 

With 6G, an always available system is emerging that ensures access to the edge and cloud infrastructure in every environment and deployment scenario. This hyperconnectivity is hardly manageable with conventional methods. The latest breakthroughs in Artificial Intelligence are therefore set to become the central driver of the sixth generation: as an integral component across all layers of the network. In this way, autonomous AI agents are to coordinate the networks in real time and anticipate potential problems. To train intelligent decision-making and to simulate scenarios, digital twins are employed. However, a challenge lies in tailoring the AI methods to the specifications in the upcoming standard.

Our main focus is on resource allocation. We study how wireless networks can automatically adapt to changing conditions and demand using AI-based predictive analytics, so that the available bandwidth, power and time slots are dynamically distributed while Quality of Service is guaranteed. With joint source and channel coding, we use AI to improve physical layer transmission efficiency and overall Quality of Experience. Another key area is neuromorphic computing technologies such as spiking neural networks, which mimic the functioning of the human brain and transmit data in the form of pulses. We explore how these neural networks can enable the energy-efficiency requirements of 6G signal transmission. One other machine learning paradigm in our repertoire is federated learning. This refers to the decentralized training of neural networks directly on the end device. In order to open this learning process for 6G, we have incorporated a federated learning architecture in our system-level simulations. 

One central promise that the 6G standard aims to fulfill is real-time capability with extremely low latencies of less than one millisecond. However, 6G can only play this card if the reliability of data flow is fully ensured, since any failed transmission would increase latency again. Especially mission-critical applications cannot afford outages or delays, as this would lead to significant production losses or safety risks. The challenge for mobile communications lies in dealing with so-called "blockages," which are physical obstacles that reflect or weaken signals, thereby severely impairing transmission quality.

Real-time and reliability come as a package deal with us. The focus is on industrial halls: we research technologies to enable reliable real-time radio even amidst metal walls and shadowing zones for mobile participants. This includes fast dynamic rerouting: if an obstacle is detected, the network can set up an automatic detour and send the data traffic via an alternative path. Another focus of our work is on sub-networks in robots and machines. With “UWIN,” we have developed a proof of concept that prototypically demonstrates how the limits of real‑time communication can be pushed within such small, local networks. Additionally, we are working on making Time-Sensitive Networking universally usable for 6G. Through prioritization and coordination, it is guaranteed that critical information always arrives at the right place at the right time. 

It is one of the most groundbreaking innovations compared with previous generations: with 6G, mobile communications finally gain eyes and ears. Signals normally used for data transfer are reflected off objects as they travel from the transmitter to the receiver. In sensing, the networks use these echoes to locate, analyze, and track objects. This means that communication signals can also serve as radar waveforms within the existing mobile infrastructure. Since base stations are already widely available, a densely meshed sensing network can take shape, extending the capabilities of traditional cameras and sensors. The avoidance of specialized radar hardware not only saves operating and maintenance costs but also facilitates potential use cases such as tracking vehicles, monitoring areas, and detecting drones in airspace.

Because communication signals were not originally designed for use as radar, they still lack the necessary granularity. Therefore, our research focuses on optimizing the waveforms in 6G so that the dual use of communication and sensing is possible without performance losses. The goal is to elevate environmental perception through the intelligent combination of mobile communications, sensors, and cameras to a new level. Coexistence should lead to cooperation: We are developing multimodal AI methods and synchronization techniques to effectively fuse data from the different sources into a holistic picture.