6G - Taking radio access technologies to the next level

With 6G standardization waiting around the corner – 3GPP will start the technology studies this fall – it is timely to take a look at the potential of 6G and the gains achievable beyond 5G. In an earlier blog post, we discussed a number of technology components we envision for 6G. In this post, we will examine the performance potential of some of them. 

Standardizing a new 6G RAT is a once-in-a-decade opportunity to significantly improve performance and allow enhancements to fundamental aspects not possible to change in 5G. 

Enhancing energy performance by making explicit changes to signals used for connection establishment is one example that we will discuss shortly. Another example is further improved resilience, an important area as society increasingly relies on mobile communication. Resilience is a wide area and spans from proactive measurements in the radio network to complementary use of satellites to resilient compute platforms. A deeper discussion on this topic is is left for another blog post.

The new 6G RAT is also envisioned to be inherently highly scalable in terms of data rates and addressed use cases while aiming for reduced system complexity and operating costs. 6G will support both high-end devices that surpass the capabilities of today's 5G and low-cost, low-end devices that can serve the needs of the massive IoT segment. Eventually, 6G is expected to replace older 4G-based LTE-M and NB-IoT technologies used today. Not only will this over time lead to simpler network management – handling one 6G network is simpler and can reduce cost compared to handling multiple networks – it also means that the general 6G benefits such as improved network energy efficiency will be available also to the IoT segment.

With this introduction in mind, let’s look into some examples of the performance potential of 6G!

Additional background to the performance examples are available in the document Detailed results and simulation assumptions. Note that different assumptions may result in different numbers, that is, the results should be seen as an illustration of the technology potential.

Energy efficiency – a key property of both networks and devices

The cost of energy constitutes a large part of an operator’s OPEX and is one of the motivations behind the ultra-lean design of 5G. In 6G, the ultra-lean design paradigm should be evolved and expanded beyond the time domain to also cover the frequency and spatial domains. 

One opportunity is to increase the periodicity of the always-on synchronization signal block (SSB), which is used for example in initial access and mobility. Going from 20 ms as used in 5G to 160 ms, there is potential to reduce the energy consumption in an idle network by up to 77 percent. 

Such changes are difficult, if not impossible, to introduce in 5G because they require changes to the fundamental connection establishment procedures, which would impact devices already in use. In a new generation, such design is much easier. By complementing the sparse always-on SSB with on-demand SSBs for connected devices when needed, the ultra-lean paradigm can be expanded into the spatial and frequency domain and result in even larger gains in some deployments. 

Long battery lifetime is high on end-users’ wish lists, underlining the fact that energy efficiency is equally important for devices. This should be accounted for in the 6G design, for example by including a wake-up signal (WUS) to activate devices when data arrives. This allows the device to put most of its circuitry to sleep to save energy, yet be highly responsive to incoming data packets. Studies have shown around 35 percent reduction in device energy consumption in the common scenario of small, bursty packets – without any negative impact on quality-of-experience for the end-user. 

Massive antenna arrays and MIMO – an essential component of 6G

MIMO and large-antenna-arrays will be an essential component of 6G and enable the use of cmWave bands to unlock new capacity. For example, adding 200 MHz of cmWave spectrum at 7 GHz to an existing low-band/mid-band macro deployment can triple the downlink capacity (+200 percent). 

The key takeaway is primarily not the gain in itself – additional spectrum typically provides gains – but the fact that the 7 GHz band can be used on existing macro-sites without the need for further densification! At higher frequency bands, it’s possible to have a larger number of antenna elements without increasing the size of the antenna installation. The beamforming gains achievable with this antenna configuration compensate for the somewhat more challenging propagation conditions. CmWave bands can provide a valuable spectrum asset to be exploited by 6G.

Massive MIMO is widely used for TDD midband and highband. Massive MIMO can also be beneficial at lower frequencies. Due to their lower attenuation and use of FDD, these bands, especially in the uplink, provide better coverage than higher bands using TDD. Thus, they are important for overall system coverage and for providing capacity to users who need that coverage. Further improving the coverage and capacity of these bands using massive MIMO makes it possible to use new services in a certain coverage area or to extend the range in which a particular service can be used. 

In one example, we looked at a network deployed in an urban area using frequency bands ranging from 700MHz to 7GHz. When we upgraded the 1800-2600MHz frequency bands with massive MIMO, the uplink cell-edge data rate more than doubled for the same traffic load. Alternatively, we could double the network capacity while maintaining the same cell-edge data rate. 

Although massive MIMO in FDD is feasible already in 5G, there are limitations. For example, the number of downlink antenna ports supported by the existing devices for the necessary channel-state reporting is a limiting factor. This is particularly true in combination with carrier aggregation.

AI will play a pivotal role in 6G and can, in selected areas, significantly improve the performance relative to conventional algorithms. It is worth emphasizing the importance of comparing with state-of-the-art implementations and not with simple, textbook-based algorithms – as is sometimes seen in various publications. 

One example where AI can shine is in MIMO channel prediction. To make the best use of the massive arrays, the base station requires accurate channel state information. Conventional schemes either rely on uplink sounding signals (SRS), which works well in the cell center but falls short in outer parts of larger cells as the signal gets weak, or on feedback from the UE – a good choice in coverage-limited scenarios but requires a downlink reference signal transmission and an uplink feedback mechanism introducing overhead. AI-based SRS receivers have the potential to outperform the two classical schemes. Read the Extended Information pdf to learn more.

Uplink performance is becoming increasingly important

Uplink performance is becoming increasingly important as we move from a world dominated by downlink data consumption into a future with extensive XR usage, uplink video traffic, and AI agents in the devices. To efficiently meet this demand, uplink control signaling and the scheduling strategy need to be revisited. 

Control signaling sent inband together with user data (that is, on L2) instead of a separate physical channel as in 5G is one interesting possibility. It will significantly simplify the overall scheduling timeline and allow for a wide range of deployment flexibility. Not only that – it will also provide a 56 percent gain in experienced bitrate for small objects in the uplink and, interestingly, at the same time a 30 percent gain in downlink performance.

The simulation in this case is a full end-to-end simulation including all protocol layers (PHY, MAC, RLC, and TCP) and therefore reflects what a user would experience in a real system. This is important, since evaluating parts of the system only, or uplink only, would not reveal the full picture.

Real cellular systems are multi-band systems with multiple frequency bands typically available at each site. Higher frequency bands, for example, 3.5 GHz TDD or the cmWave range, offer bandwidths in the order of hundreds of MHz. Lower frequency bands, for example, sub-GHz FDD bands, can offer better coverage due to more favorable propagation conditions although the available transmission bandwidth is smaller. 

In 5G, a device typically camps on the best downlink frequency and uses the same frequency band also for its uplink traffic. However, given the difference in transmit power and antenna configurations, a device in a location with less favorable radio conditions could sometimes be better served if it could transmit in a lower frequency band but receive data on wider bandwidth in a higher frequency band.

Uplink-downlink decoupling, where a device can be independently and instantaneously assigned the best frequency bands for uplink and downlink, can provide up to 10x improvement in uplink data rates at low system load!

Beyond communication – positioning and sensing

Not only will 6G improve communication performance, it will also expand into areas beyond communication. Positioning is an example of a beyond-communication service available in 5G. With the support of AI, accuracy can be improved. In a real outdoor setup, with mixed line-of-sight and non line-of-sight propagation and a single base station, the mean error can be reduced from 7.1 m to 5.5 m by employing AI, an accuracy improvement of 23 percent. In very problematic non-line-of-sight scenarios, a hundred-fold improvement in positioning accuracy from the use of AI has been found – an example where AI can be a true game-changer. 

Integrated sensing and communication (ISAC) is another example, new to 6G, of a beyond-communication service.. simulations indicate that by reusing existing sites, radios, and frequency bands, drone detection in central London is possible with an average accuracy of 2 m. In a live test network with realistic measurements, an average horizontal error of 2 – 3 m has been observed, which is in line with the simulations. It’s reasonable to conclude that ISAC has a promising and realistic potential, for example, to enforce no-fly zones for drones.

Promising potential in 6G

In this blog post, we have illustrated some of the technology potential of 6G radio access. The final 6G specifications are still a few years into the future, details remain to be settled, and gains may depend on the scenario studied. Nevertheless, it is clear that there is great potential in a new generation of 6G wireless access for the next decade!

Download the Detailed results and simulation assumptions pdf for more about the studies.

Read more

Visit the 6G website
Blog post 6G RAN – key building blocks for new 6G radio access networks
Blog post 6G standardization: The technology realization step begins
Blog post 6G: Are we ambitious enough?