Are Reconfigurable Intelligent Surfaces Ready for 6G?

Prof. Emil Björnson has featured in many posts across our different blogs over the years. In fact, one of our earliest posts on Reconfigurable Intelligent Surfaces, or RIS, was back in December 2020, when we featured his explanation of how Intelligent Reflecting Surfaces could be used in B5G and 6G systems. Five years later, Prof. Emil Björnson and Prof. Erik G. Larsson have returned to the topic in the Wireless Future Podcast to ask a very timely question: are reconfigurable intelligent surfaces ready for the world?

RIS has been one of those technologies that attracted huge interest in the early 6G research phase. It sounded almost magical: instead of accepting the radio environment as fixed, why not make parts of it programmable? A wall, panel or surface could be engineered so that when radio waves hit it, the surface can reflect or redirect them in a controlled way. In simple terms, RIS can behave like an intelligent mirror for radio waves.

The basic idea is easy to explain. In wireless communications, we normally depend on whatever propagation paths are available between transmitter and receiver. Sometimes there is a strong line-of-sight path. Sometimes the signal reaches the receiver after reflecting off buildings, walls or other objects. Sometimes a user is blocked by a corner, building, hill or other obstacle. RIS offers the possibility of creating an additional controlled propagation path, sometimes described as a “virtual line of sight”. The base station has line of sight to the surface, the surface has line of sight to the user, and the signal is reflected in the desired direction.

That is the promise. The podcast discussion is interesting because it goes beyond the hype and looks at what has actually happened since RIS became fashionable in wireless research around 2019 and 2020. The conclusion is not that RIS is useless, nor that it is ready to transform every mobile network. The more balanced answer is that RIS has become much better understood, prototypes now exist, experiments are happening, but the practical use cases are narrower and more demanding than early research papers sometimes suggested.

One of the most convincing use cases discussed is fixed wireless access. This makes sense because the transmitter, surface and receiver are mostly fixed. If a base station cannot directly reach a group of homes, an RIS could be placed at a suitable location, for example on a building, pole or elevated point, to redirect the signal towards those homes. Since the geometry changes slowly, the RIS configuration does not need to be updated at the same speed as in a highly mobile scenario. This makes the control problem much easier.

This is important because configuring an RIS is not trivial. Emil describes a lab prototype operating at 28 GHz, with 32 by 32 controllable elements, giving 1,024 elements in total. Each element can switch between two phase states. In theory, that creates an enormous number of possible configurations. In practice, one does not search through all combinations. Instead, the surface is configured based on knowledge of where the signal is coming from and where it should go. If the geometry is dominated by line-of-sight paths, the problem becomes much more manageable.

Another relevant use case is improving capacity or coverage in specific parts of a mobile network. For example, an RIS could help redirect signals into a blocked street, a park, an indoor area, a tunnel, a factory or another location where coverage is weak. The biggest benefits appear when the RIS has line of sight to both the base station and the target area. In other words, RIS is not a magic cure for every coverage problem. It is most useful where the deployment can be planned and where the surface can create a strong and predictable additional path.

This is also why RIS becomes more interesting at higher frequencies. At lower frequencies, radio signals bend, scatter and penetrate better. At higher frequencies, especially millimetre wave and potentially sub-THz bands, propagation becomes more like light: either the path is available, or it is blocked. In those situations, a controlled reflecting surface can make a much bigger difference. The paper discussed in the podcast, “Reconfigurable Intelligent Surfaces in Upper Mid-Band 6G Networks: Gain or Pain?”, focuses on the upper mid-band, roughly 7 to 24 GHz, because this range is often discussed as a likely candidate for early 6G deployments.

A key lesson from the discussion is that RIS gains are highly scenario dependent. Emil mentions simulations where users could gain one or two additional bits per second per hertz, which could translate into a substantial capacity improvement. But this depends strongly on the propagation conditions, the deployment geometry, and whether there are already useful static paths between transmitter and receiver. One criticism of some earlier RIS research is that it assumed the only possible path was via the RIS, completely ignoring weaker but still existing direct or reflected paths. That can make the RIS look more beneficial than it would be in the real world.

The podcast also makes a useful comparison with network-controlled repeaters. At lower frequencies, a repeater may be simpler, cheaper and more practical than an RIS. A repeater receives a signal, amplifies it and retransmits it. RIS, especially passive RIS, does not amplify in the same way; it redirects the incident signal. That can reduce power consumption, but it also means the deployment must be carefully planned. At high frequencies, however, repeaters also need directional antenna arrays and beam control, which brings them closer to some of the same challenges faced by RIS.

The practical measurements described by Emil are especially valuable because they show both the promise and the limitations. His lab experiments demonstrated that a RIS can create a virtual line-of-sight path and perform beam focusing in the near field. At the same time, the experiments revealed practical issues. Other objects, such as a laptop cover, can also reflect signals and affect measurements. Large reflection angles may not work as cleanly as expected. The element pattern, quantisation of phase shifts, coupling between elements and control software all matter. This is exactly the kind of detail that separates a good theoretical idea from a deployable technology.

The discussion then moves into newer RIS variants. Active RIS adds amplification at the elements, making it closer to a repeater, but this also brings transmitter noise, power consumption and complexity. Beyond-Diagonal RIS introduces controlled coupling between elements, which can in theory support more advanced transformations, including handling multiple incoming and outgoing directions. The challenge is that the hardware becomes more complex, losses may increase, and channel estimation becomes much harder.

STAR-RIS stands for simultaneously transmitting and reflecting RIS. Here, “transmitting” does not mean the surface is acting like a normal transmitter. It means the wave can pass through the surface to the other side, while some energy may also be reflected. This could be useful for thin walls or partitions between rooms, but the deployment scenarios are more constrained than for ordinary reflecting surfaces.

Stacked Intelligent Metasurfaces are another emerging idea. Multiple transmissive metasurfaces are placed in layers so that the radio wave is gradually processed as it propagates through them. This creates a kind of wave-based analogue computing structure. The concept is fascinating, especially for future radio processing, but the podcast rightly points out that much of this work is still theoretical. In practice, every layer introduces loss, reflections and implementation imperfections. More layers will not automatically mean better performance.

So, are RIS ready for the world? The answer seems to be: ready for serious experimentation and selected niche deployments, but not yet ready to become a mainstream 6G feature everywhere. The most realistic early opportunities may be in fixed wireless access, indoor coverage enhancement, factories, mines, tunnels, high-frequency backhaul or fronthaul, and other controlled environments where the geometry is known and stable. Passive or semi-passive surfaces may also find use where a fixed reflection pattern is sufficient, without full real-time reconfiguration.

For 6G, this is a useful reality check. RIS remains an exciting technology because it changes how we think about the radio environment. Instead of only designing better transmitters and receivers, we can also think about engineering the space between them. But the technology must compete with simpler alternatives, including better site planning, repeaters, relays, distributed antennas and more base stations. The question is not whether RIS can work. The question is where it works well enough, cheaply enough and reliably enough to justify deployment.

That is why this Wireless Future Podcast episode is worth watching. It captures the maturity of the RIS discussion in 2026. The hype has cooled, the research has become more practical, hardware prototypes are available, and the strongest use cases are becoming clearer. RIS may not be the universal 6G game-changer that some early presentations suggested, but in the right place, at the right frequency, and with the right deployment model, it could still become an important tool for shaping future wireless networks.

The full discussion is embedded below:

Related Posts: 

Posted by on

Comments

Post a Comment