Introduction

6G technology is the next generation of wireless communication that is expected to offer unprecedented speed, capacity, and latency. It will build on the foundation of 5G and introduce new capabilities that will enable novel applications and services. 5G networks use a combination of sub-6 GHz and millimeter wave frequencies, as well as advanced technologies such as massive MIMO, beamforming, and network slicing, to deliver high speeds and ultra-reliable low-latency communication. However, 5G still faces some challenges, such as spectrum scarcity, energy efficiency, and coverage. 6G aims to overcome these challenges and achieve even higher performance goals.

According to some estimates, 6G will be able to deliver terabit speeds, 1-microsecond latency, and far greater capacity than 5G. To achieve this, 6G will use a number of enabling technologies such as higher frequency bands including terahertz and sub-terahertz, intelligent reflecting surfaces, artificial intelligence, and novel waveforms and physical layer techniques. 6G will also leverage satellite networks and non-terrestrial platforms to provide ubiquitous coverage.

The development of 6G is still in its early stages, but some milestones have been set by international organizations and industry players. The ITU has launched the IMT-2030 vision project to define the requirements and roadmap for 6G. The 3GPP has started the study on beyond-5G systems and plans to release the first standard for 6G by 2028. Several countries have also launched research initiatives and testbeds for 6G.

The potential applications of 6G are diverse and far-reaching. 6G will also enable new paradigms of communication that require new metrics and quality of service parameters to ensure user satisfaction and system efficiency.

Given the complexity of bringing 6G to market, researchers and wireless engineers are going to need to rigorously simulate, test, and experiment with different software tools. Programs such as MATLAB^® are going to be indispensable in tackling the most difficult research questions posed by 6G. This white paper will cover some of the key tools already available to get started on building the next generation of wireless technology.

6G Use Cases and Requirements

5G adoption is up and running around the world. However, the continuous convergence of the physical and the virtual world across many domains will further magnify performance requirements and push 5G to its limits in the long run. Therefore, the next-generation (6G) wireless systems will need to reach unprecedented service quality levels capable of satisfying a whole new class of applications and services for 2030 and beyond.

Beyond improving on existing 5G use cases, some researchers envision that 6G will need to handle extremely demanding applications such as holographic communications, extended virtual reality (XR), massive digital twinning, and ultra-large-scale Internet of Things (IoT). Such use cases will generate massive amounts of data, require ultra-high bitrates at precise locations, and provide network efficiency clearly superior to what 5G can offer. Such applications will also require intelligence capabilities beyond 5G to enable real-time decision-making on high volumes of data.

6G applications can be classified into different categories based on high-level functional and performance requirements. The focus of this white paper is on four categories:

• Networked-enabled robotic and autonomous systems: Applications where systems can perceive their surroundings using sensors and interact with humans in natural ways and make decisions necessary to assist or support a set of tasks. Applications include online cooperative operation among service robots and digital twins for manufacturing.
• Multisensory extended reality: Application for advanced virtual reality (VR) and augmented reality (AR) bringing highly immersive experiences with haptics, visuals, and audio, adapted to the environment. Applications include mixed-reality co-design and mixed-reality telepresence.
• Distributed sensing and communications: Use cases with massive sensor and data collection networks. Applications include in-body networks and immersive smart cities.
• Sustainable development and inclusive communication: Use cases in this category focus on reducing inequalities and achieving digital inclusion by ensuring global access to digital services. This includes remote medical services, expanded digital access, and additional educational resources in areas historically difficult to reach with wireless internet.

6G Research Ecosystem

Although 6G standardization work is not expected to begin before 2025, several initiatives have been started around the world to conceptualize what 6G will look like. In what follows, this white paper will go through a sampling of initiatives and activities worldwide to create an overview of the 6G research ecosystem.

Internationally, the ITU Radiocommunication Sector (ITU-R) of the International Telecommunication Union (ITU) has tasked a working party (WP 5D) with crafting the vision for mobile communications beyond 2030 in the form of a recommendation.

In North America, the Next G Alliance aims at establishing North American leadership in 6G research and development.

In Europe, the 6G charge is spearheaded by Smart Networks and Services Joint Undertaking (SNS JU). Moreover, several EU-funded flagship research projects on 6G have been started.

Similarly in Asia, several initiatives have been recently launched to define 6G vision and enabling technologies.

Enabling Technologies

Among the various 6G initiatives around the world are some common technologies envisioned to bring 6G concepts into reality. Experts agree on these key technologies for 6G:

• Artificial Intelligence
• Joint Communications and Sensing
• Reconfigurable Intelligent Surfaces
• Non-terrestrial Networks (NTNs)
• Physical Layer Design
• Extreme Data Rates and Higher Frequencies

In the following sections, this white paper goes through those common denominators and provides insights on how you can use MATLAB to tackle the most difficult research questions of 6G.

Deep Learning and AI

Artificial intelligence (AI) is already used in 5G and is being considered for a wide range of use cases within 6G research. An AI workflow requires architecting deep neural nets and generating vast amounts of data for training these nets, including GPU support for efficient training, and MATLAB has all these capabilities. Use cases include:

• Beamforming design
• Adaptive channel estimation
• Data-driven channel decoding
• Compensation for hardware impairments

In 3GPP Release 18, particular focus is put on three AI areas:

• Channel state information (CSI) feedback compression: Neural nets could be utilized to compress the CSI fed back from the receiver to the transmitter.
• Beam management: An exhaustive search over all beam pairs in a massive MIMO system can become numerically prohibitive. An alternative is to utilize AI to reduce the search space to a smaller set of beam pairs.
• Positioning: Accurate positioning enables multiple applications but is technically challenging. AI has the potential to increase positioning accuracy.

Using MATLAB for Deep Learning and AI

MATLAB supports the full deep learning/AI workflow, from initial idea to a trained neural network running on an embedded device.

MATLAB toolboxes help researchers across the entire workflow, from data preparation to deployment.

MATLAB and related toolboxes let you design, train, test, and deploy deep neural nets for a variety of applications. MATLAB comes with a large set of executable demos in AI for wireless applications:

Joint Communications and Sensing

One of the potential features of future 6G networks is the ability to utilize the radio spectrum for sensing as well as for communications. Joint communications and sensing refer to a new paradigm in which the radio hardware and software could perform both tasks of sensing and communications. Potential use cases include traffic monitoring and localization of passive objects; environmental monitoring and human activity/presence detection; and fall detection and blood glucose monitoring.

Joint communications and sensing can be divided into two approaches:

• The same radio spectrum is used for sensing and for communications. This entails adding new signal processing for sensing at the receiver, but no changes are needed to the communication functionality. Sharing the spectrum between sensing and communications poses interesting challenges, since there has to be a tradeoff in the waveform design between sensing capability and communications performance. For example, the Cramér-Rao lower bound (CRLB) could be a valid metric to be used for sensing, whereas capacity is a better metric to be used for communications.
• Different parts of the radio spectrum are used for sensing and for communications. Dedicated hardware could be used for the purpose of sensing. This approach then becomes a matter of how to share the available radio resources over time, frequency, and the spatial dimension.

Using MATLAB for Joint Communications and Sensing

Combining Communications Toolbox™ and Radar Toolbox, you can easily conduct experiments in joint communications and Sensing. Communications Toolbox has building blocks to set up the communications signal processing chains, whereas Radar Toolbox contains all the textbook algorithms you will need for the sensing part.

Positioning and localization are central concepts in many sensing applications. MATLAB offers many in-depth examples in these areas. A system that integrates communications with sensing needs to make a tradeoff between the two kinds of waveforms. MATLAB lets you study this tradeoff. Micro-Doppler signature detection is a technique that enables many of the use cases studied with joint communications and sensing. MATLAB lets you study micro-Doppler signature classification, e.g., using deep learning techniques.

Reconfigurable Intelligent Surfaces

Reconfigurable intelligent surfaces (RIS) are paradigm shifting techniques that allow the manipulation of the wireless channel to provide ultra-reliable coverage and superior communication quality. A classical wireless system considers the propagation environment as a given. Thus, its goal is to optimize the communication performance by adapting its transmission schemes and parameters such that the impairments of the “given” channel are overcome.

RIS are planar surfaces consisting of reflecting elements that can independently and passively influence the phase of the signal they reflect. Through programmable elements, RIS can reconfigure the wireless channel by tuning the phase shifts of a large number of reflector elements on a surface or antenna array. This will give the communication system active control over the characteristics of the radio environment and allow it to eliminate or enhance certain signal propagation directions as well as suppress interference.

The research community is already working on a set of challenges and research problems that need to be addressed to bring RIS from theory into practice:

• The design of reflective surfaces with large numbers of elements to find suitable scenario-specific configurations for controllable elements in a timely manner while minimizing signaling overhead
• Accurate estimation of the wireless channel between RIS and transmitters or receivers and CSI acquisition considering the large number of reflective elements and highly dynamic scenarios as RIS get mounted on UAVs
• Design and optimization of robust beamforming considering the imperfect nature of the acquired CSI from RIS systems

Using MATLAB for Reconfigurable Intelligent Surfaces

Using Phased Array System Toolbox™, Antenna Toolbox™, and Optimization Toolbox™, you can model and design scattering surfaces and dynamically change their characteristics. Additionally, MATLAB allows you to:

• Model reflective surfaces and elements with an extensive catalog of elements including dipole, monopole, patch, spiral, fractal, and horn antennas.
• Design optimization algorithms to optimally control the different elements of the reflecting surfaces.
• Flexibly design antenna arrays such as linear, rectangular, circular, conformal arrays, and custom array designs to explore the design space for RIS.
• Accurately model 3D propagation environments using ray tracing to compute multipath propagation paths while taking into consideration ITU Permittivity and conductivity values of common materials.
• Model multipath propagation–scattering MIMO channels to model reflections from multiple scatterers toward a receiving array. The model takes range-dependent time delay, gain, Doppler shift, phase change, and atmospheric loss due to gases, rain, fog, and clouds into consideration.

Non-Terrestrial Networks

Using MATLAB for Non-Terrestrial Networks

Existing 5G NTN link models can be used as a starting point to investigate required improvements and better algorithms for 6G. With the help of the 5G Toolbox™ and the Satellite Communication Toolbox, MATLAB has all the tools in place to accelerate your NTN research by allowing you to: 

• Flexibly build, model, and visualize complete satellite scenarios consisting of satellites, constellations, and ground stations and the interactions between them.
• Perform multidomain modeling with satellites and aircrafts in the same scenario.
• Model multi-hop satellite links between ground stations and perform access analysis between constellations and ground stations.
• Model Doppler frequency and latency between satellites and ground stations.
• Use standards-based satellite propagation models to model propagation losses and cross-polarization as defined in the ITU-R P.618 recommendation as well as the land mobile-satellite (LMS) communication channel as defined in the ITU-R P.681-11 Section 6.2 recommendation.
• Explore new waveforms, modulation schemes, and error correction techniques with an extensive communications library of functionalities.
• Design, model, and test beamforming systems to meet beam width requirements for NTNs.
• Design antennas and antenna arrays capable of realizing beam width and efficiency requirements typically used in NTN.

Extreme Data Rates and Higher Frequencies

An aspirational goal in 6G is to provide data rates up to hundreds of Gbps. Multiple new challenges are associated with extremely high data rates, of which some are related to increased power consumption and higher carrier frequencies:

• A signal bandwidth in the order of tens of GHz will be required to attain extreme data rates, even if the spectral efficiency is high. In turn, this means that the carrier frequency must be in the upper mmWave region (>100 GHz). When it comes to RF propagation, the main challenge in higher frequencies is the high attenuation. To correctly represent these limitations, new channels will be needed for upper mmWave and sub-THz bands. Basing such channel models on stochastic modeling, as is standard practice for lower frequencies, is challenging for higher frequency bands. Channel models based on ray tracing have provided good prediction capabilities at 60 GHz, and similar capabilities are also expected at higher frequencies. Ray tracing models lend themselves well to beamforming, which is a crucial technique to overcome the range problem.
• For data converters, power consumption increases approximately linearly with the sampling frequency, but exponentially with the bit resolution. New challenges imposed by an increased power consumption due to higher bandwidths may require redesigns of digital-to-analog converters (DAC) and analog-to-digital converters (ADC), e.g., by reducing the bit resolution.
• Data rates will be much higher than the clock rate for the DSP circuits, requiring novel DSP algorithms designs to process massively parallel data streams.

Using MATLAB for Extreme Data Rates and Higher Frequencies

MATLAB has built-in functionality for ray tracing. On top of this, the tools have built-in functionality to add losses due to rain, terrain diffraction, refraction through the atmosphere, tropospheric scatter, and atmospheric absorption (for example, see CDL Channel Model Customization with Ray Tracing and Indoor MIMO-OFDM Communication Link Using Ray Tracing).

MATLAB lets you explore and modify the architecture of data converters to a high degree of accuracy.

MATLAB has ready-made IP blocks that process data in parallel, hence attaining an effective data rate that is much higher than the clock rate. Simulink^® models using such blocks can be deployed to and run in real time on FPGA platforms.

Physical Layer Design

An updated physical layer design may consist of a new frame structure, new waveforms, and novel channel coding techniques. Waveform design for 6G comes with several challenges. The available link budget will be reduced by limitation in the peak output PA power at higher frequencies, which favors waveform candidates with low envelope variations. At extremely high data rates, the analog-to-digital conversion is expected to be a main contributor to the system’s power consumption, which favors energy-efficient waveforms. The following are examples of waveform candidates being considered in 6G:

• CP-OFDM waveforms build on a long tradition from 4G and 5G but have the drawback of having a large peak-to-average power ratio (PAPR).
• Zero-crossing modulation (ZXM) achieves high energy efficiency by reducing the amplitude resolution.
• DFTS-OFDM reduces the PAPR compared to CP-OFDM at the expense of additional signal processing.

Using MATLAB for Physical Layer Design

Communications Toolbox and 5G Toolbox allow you to explore different technologies, starting from existing 5G models and using various channel models to explore performance in different frequency bands.

MATLAB lets you explore new coding schemes based on, for example, NR LDPC and polar codes.

Conclusion

6G wireless technology offers an exciting future for wireless engineers and researchers. Over the next decade, the use cases and technologies discussed in this white paper will become increasingly prominent in the field of wireless communications.

For those looking to learn more about MATLAB and 6G, please review the recommended next steps and resources below.