6G Line-Of-Sight Repeaters, Dots, And Reflections

6G will open the door to ultra-reliable, low-latency communications, extended broadband, and machine communications, but its rapid signal attenuation places some sharp limits on where and how it can be used, and requires some expensive options to overcome those limitations.

Applications include lifelike virtual reality for home and work use, highly interactive smart homes and cities, and autonomous vehicles responding to each other. It also can be used for infrastructure such as traffic lights and industry 4.0 advancements, including responsive robotics, as well as many things we can’t yet imagine.

As with 5G, 6G comes in two flavors, sub-THz and THz systems. But unlike sub-6GHz 5G, both versions of 6G are aimed at extreme data rates of 100 Gbits/s and come with constraints, including:

• Challenging propagation conditions;
• Limited availability of commercial radio frequency components, and
• The necessity to operate within relatively small coverage areas.

“The advantages with higher frequency are that you get much higher bandwidth for very short communication between your laptop or your monitor or your television,” said Andy Heinig, head of department for efficient electronics at Fraunhofer IIS’s Engineering of Adaptive Systems Division. “But it’s only two meters, with nothing in between. You can’t do it between two rooms. What are the consequences? You need more specific protocols for different applications, such as in-car communication or in-airplane communication, where you can use this high bandwidth. You have to know what you want to reach, and it is limited.”

6G’s higher frequencies mean that antennas on cloud-connected base stations must be able to see the antennas on edge devices. “6G has very limited ability to go through buildings and concrete, so the normal antennas that we have now are not good enough,” said Marc Swinnen, director of product marketing at Ansys, now part of Synopsys. “You need many more antennas to peek around corners and make sure you get in between the buildings. You have lots and lots of edge stations that relay the signal around the corner, pick up the signal, and send it back to the main antenna and do the backhaul.”

Integrating all these components is a challenge, considering private 5G networks are expected to deliver practically limitless bandwidth anywhere in a building or site. However, there are several ways to boost coverage.

First, more small-cell radios can be used. “There are some companies out there that are building super cells, where giant individual cells can cooperate through smart pre-coding to act like a logical supercell so the small cells don’t interfere with each other electromagnetically. That way you can increase your coverage and plot down radios anywhere you have weak coverage and not worry about the interference patterns of the cell interfering with some other radio,” said Ron Squiers, solution networking specialist at Siemens Digital Industries Software. “Barring that, you’ve got to map out and understand the coverage required in the factory and what the line-of-sight issues are, the non-line-of-sight issues for diversity and reflection and absorption, and all that has to be modeled and understood.”

Another option is to install multiple repeaters to amplify and retransmit signals from the base station. For example, Ericsson has a repeater concept called the dot, which is like a local and private 5G wireless access point. “This is a turnkey solution,” said Shawn Carpenter, program director for 5G/6G at Ansys. “You get a bunch of them, and just put them out around your ceiling, across your shop floor, across your factory, across your warehouse, and then you would work with the company to determine the right density of these access points to put up, which is going to depend on you as a user. What do you need to do? How much throughput do you think you’re going to need?”

Repeaters are the best bet in next-gen factories, but companies must determine how to connect them. “You probably have lots of video inspection stuff going on, and you’ve got a crucial command control for manufacturing equipment that has to be super reliable,” said Carpenter. “You’re going to need a certain densification of the access points. You probably have to connect those access points with a fiber optic backbone. Or, maybe you want a wireless backbone, because you don’t have the option or the environment’s too hostile to lay that stuff down, like in a steel factory. It’s hard to answer the question without doing some analysis.”

Private networks are so complex that companies are building up engineering and services wings to address this. “What kind of access do you need? What kind of bandwidth requirements are you going to have? How far apart should we put these things, and where should we put them?” said Carpenter. “It involves getting into the architecture of your structures. It gets into how the outdoor campus of your facility is set up, or a whole university. You have to customize it, but they’re trying to create the off-the-shelf parts — or the components that you could put together and put [the repeaters] in the right place, and it will just all work together and talk to one another through a control node.”

Whether companies can afford enough repeaters comes down to economics. It depends on how the service providers, whether they are the traditional telcos or other players in the space, are going to charge the users, said Mung Chiang, president of Purdue University and recipient of the 2025 IEEE Founders Medal for leadership and research in communication networks and their applications. “If the users are not human, but owners of devices and machines, how ready and willing are they to be charged? It depends on the data pricing structure and the economic model. Then we’ll see under what circumstances it would be affordable to have a very dense deployment so that you have line-of-sight. Not all deployment scenarios will be in a position to afford it. Some will, both residential and consumer type, as well as the machine industry. In the end, it comes down to how you are going to charge for it. How much money can you get back so that you can afford to pay for the spectrum, buy the capital equipment, run the operational expenses, maintain backward compatibility with all the other things that you already have?”

Fig. 1: Sub-THz technology can be broadly clustered into functional enablers of access, backhaul, and sensing. Source: Purdue University 6G Global Roadmap Taxonomy Report

With 5G/6G networks covering areas as large as seaports, mobile traffic is expected to grow exponentially, but energy consumption should not, according to Infineon. “5G, therefore, requires a new network dimensioning to optimize energy consumption. This includes efficient energy-saving solutions for an entire site and the entire network. Traditional radio base stations consume 80% of the energy, although they spend most of the time inactive or without transmitting. To achieve the lowest energy consumption for sites and the entire 5G network, operators have to pay attention to the efficiency of individual components, but also to the entire link and the whole site.”

Finding and filling coverage holes in factories and cities
In a factory setting, private 6G networks will enable robots and co-bots to respond and react to each other in real time, which will improve productivity, as well as safety when they are working near humans. It also will help with automated guided vehicles in the factory.

“These solutions require digital twins to be able to model the factory, to account for things like absorption coefficients into factory walls, line of sight issues across the factory floor, basically geo-mapping the entire factory to optimally place radios within that environment,” said Siemens’ Squiers. “The tools enable companies to model the complete design environment before the factory gets built, and this is a good way of lowering the cost for customers in terms of figuring out form factors. Fit and function need to be in the factory to sustain the communications they need across it.”

Validation is a key step in all usage scenarios. “You can validate everything virtually, including all the features needed for edge computing like V2X and V2V or V2I [vehicle to infrastructure], and if you talk about 5G and 6G — all the radio — an O-RAN sender can be used for interoperability testing,” said Anup Shah, director of product management at Siemens Digital Industries Software. “For industrial applications and factory automation, everything needs to be validated pre-silicon. You want to model the entire factory environment in emulation.”

Digital twins also can be used to model signal behavior and simulate the RF propagation through an entire city. For example, Ansys used NVIDIA’s Aerial Omniverse Digital Twin platform to model San Jose. “Through simulation, we can predict where the radiation is going to go, or where it’s shadowed. Then we can enable the designer to program their receivers and transmitters to leverage the different signal paths that you might get if you have to bounce your signal two or three times,” said Ansys’ Carpenter. “The objective is to create a digital twin for a complete network in 6G that would be set up in a city. We use Ansys physics to provide a very accurate model of the antenna or the array on both ends of the channel, as well as an accurate model of the city. Then we predict how signals are going to bounce, so that you could then decide where to push the energy.”

A 6G subscriber could be a person with a phone, a vehicle, an IoT device, or anything that’s interacting with the base station or with the communications infrastructure. In cases where there is no line-of-sight visibility to a subscriber, 6G networks can take advantage of reflections to send energy in their direction, Carpenter explained. “This real-time digital twin capability for networks could allow the developer to run the digital twin in the base station and have it self-heal. It could predict where the subscriber is going next, so the antenna can be tuned forward to throw the energy where it needs to go based on where that subscriber is headed.”

Additionally, intelligent reflecting surfaces (IRS), which are similar to naturally occurring reflections, could be implemented. “These are passive panels that can be set up, but their job is to act as a special tunable reflector, a reflection surface for the RF energy that hits the wall or panel,” said Carpenter. “You can tune these things so you can control what direction they reflect in. They use metamaterials, where we can actually shape the way something reflects. It’s not like a pure mirror surface, but you can get some wacky reflection directions off this if you control it well. That will probably be part of the approach for being able to push signal content to areas that you can’t quite reach well with line of sight.”

Signal behavior in various weather conditions also can be predicted. “You can simulate how the signal will evolve depending on your position versus all these edge stations and base stations at these frequencies, and also how things like rain, fog, and mist affect signals,” said Carpenter. “They also absorb the radio waves. Radio doesn’t go through water, or at least not far. So when it rains heavily, the signals get weakened, and that can be modeled, too.”

Setting up a network with Wi-Fi 7
To install a private 5G/6G wireless network, companies have to set up their own equipment, and probably want to work in conjunction with the Wi-Fi protocols that people are using for the devices they operate with throughout the enterprise. “When you look at the speeds we get from Wi-Fi 6e and Wi-Fi 7, there’s a lot of throughput there, and you can handle quite a large number of devices, as there are mesh networking protocols for these unlicensed band capabilities,” Carpenter said.

It remains to be seen whether companies can make unlicensed band Wi-Fi protocols work together with the licensed band protocols. “Ultimately, you have to have some license band work, because you want everyone’s mobile phones to work and to be able to communicate,” he said. “Ideal target customers for this technology are chemical refineries, large warehouses, and factories in conjunction with large engineering organizations, as they need connectivity from top to bottom throughout the enterprise.”

Fig. 2: Wi-Fi 7 versus previous generations. Source: Synaptics 

Like other network standards, Wi-Fi 7 will be backward compatible. However, even though older devices won’t get the full benefit of Wi-Fi 7 support, they will get faster reaction time because the wireless medium is handled much more intelligently.

Furthermore, Wi-Fi in a private network can be used for much more than connecting two points and pushing data between them. “We are taking a broader view,” said Ananda Roy, senior product manager for low-power Edge AI at Synaptics. “Why don’t we use the wireless signals, be it Wi-Fi, Bluetooth, or Zigbee, to know more about the environment. You can leverage that and marry that with an edge AI IoT SoC to use the wireless signal for Wi-Fi sensing. As the Wi-Fi signals bounce off walls and floors, why don’t we leverage that with AI to know if there’s a person in the room? Think of a security camera, a security panel, or a thermostat. If it knows about the environment without having to have another sensing device, like a radar or a PIR sensor, it saves costs to build that system, yet makes it intelligent.”

As for residential users who like to build their own PCs, could they build their own private network with a base station to solve line-of-sight challenges in their home? “Maybe you could,” said Ansys’ Carpenter. “But part of what you’re limited by is the bandwidth. And what’s the unifying architecture that’s going to enable the different architectures to talk to one another? I’ve built my own computer several times, but I wouldn’t want to touch that. That’s a whole new level of integration and speed.”

Connecting 6G to edge devices
5G/6G is a power hog in a small ear pod or a low footprint pair of AR glasses, noted Frederic Nabki, CTO of Spark Microsystems, which creates ultra-wide broadband devices to solve this problem. “As much as it’s a high-performing system, it’s extremely power hungry, and frankly not very cheap. So you want to put it only where you truly need it, and have a fairly sizable battery, such as in a smartphone.”

For this reason, different communications protocols may be used for each hop along the fog continuum, from the cloud to fog nodes such as routers, switchers, or servers, to the last hop between base stations and edge devices.

The concept of the last hop is morphing into new terrains, said Purdue’s Chiang. “We think of the phone as the edge, but to a little device even further out from the source of data and action, the phone is like the cloud in terms of this capability — but the distance is much shorter. They are usually direct line-of-sight. That’s good, but the battery and user experience form factor constraints are limiting. I don’t have a crystal ball on exactly what bandwidth and wireless protocol technology would be useful. But just like Wi-Fi back in the mid-’90s, people say, ‘Hey, if I can use unlicensed band first of all, and then primarily stationary or very low speed, the handoff is not an issue. The cell passing to each other is not an issue. I just don’t want the wire attached to it. Maybe there’s a new protocol. The Mac is fundamentally different, and the physical layer and the waveform choices are also different, and I can afford to be unlicensed, because of those choices.’ The technology and the spectrum availability and the use case are intertwined. So could there be a room for a solution that’s between Bluetooth and Wi-Fi?”

All of the latest edge devices, such as IoT, drones, XR robotics, glasses, smartphones, and PCs have AI capability, and the availability of data with 6G allows them to have the low latency and higher data throughput that will be needed for future applications, noted Hezi Saar, executive director product management for mobile, automotive and consumer IP at Synopsys, and chair of the MIPI Alliance. “Sometimes this is a hybrid connectivity between the cloud and the device itself. You can say every device has some AI capability from the connectivity perspective, and we are seeing movement to what I would call faster connectivity. But the first thing you need is access to the data. You need to store the LLM, so you need to have gigabytes, quick read access, and a lot of read access. That’s JEDEC UFS relying on the VP specification to have this high capacity, high bandwidth, low latency access, and we also see it on the DDR connectivity. These are two must-haves to enable this low-latency capable device that when you ask for something, you get a response quickly. Because of 6G, we’re seeing faster connectivity between the SoC and the modem. Sometimes it’s a monolithic modem and SoC, depending on which vendor you’re talking to. Sometimes it’s two separate chips.”

Better connectivity and AI also make it easier for DSPs to get more accurate measurements. “For example, there is a partner that we work with that uses a wearable for GPS, and they could get more accurate local tracking if there’s dead reckoning,” said Steve Tateosian, senior vice president of IoT, consumer and industrial MCUs at Infineon. “If the GPS is not communicating for some reason using these other sensors, then the edge AI device is used to try to track what happened while we’re not talking to the satellites. Using machine learning or AI to take over that dead reckoning can improve accuracy. Similarly, for other algorithms, more sophisticated things can be done — for example, to know if you’re doing a push-up versus an overhead lift. It’s very difficult to do that with DSP. It gets a little more capable with AI.”

Conclusion
While 6G’s line-of-sight constraints will create challenges for network providers, the advantages of ultra-low latency and high bandwidth will mean that certain companies and cities go full steam ahead with installing the necessary infrastructure. For others, economics and the physical difficulty of certain sites will mean they stick with 5G and other networks.

Related Reading
6G Rollout Will Be A Patchwork At First
Spectrum allocation, infrastructure development, and varying use cases will affect when and where this technology rolls out.
Enabling Secure 5G Standalone (SA) Core Deployments
Challenges and testing strategies for a resilient network.
Innovating For 6G
Emulation is crucial for testing the performance of 6G systems in real-time channels and networks.
The Digital Twin Technology Applied To 6G Communication
6G digital twin: evolution, applications and architectural integration.