Programming the Wireless Propagation Environment

361 concepts230 papers1 proposal

For decades, wireless communication design has operated under a fundamental constraint: the physical environment is a passive, uncontrollable black box. Engineers have spent generations building increasingly sophisticated transmitters and receivers to cope with the harsh realities of fading, shadowing, and blockages. Whether through complex equalization, massive MIMO beamforming, or robust error-correction codes, the goal has always been to adapt to a random channel.

A new paradigm shifts this foundational assumption. By embedding software-controlled metasurfaces into the physical environment, we can transform the wireless channel from a static obstacle into an active, programmable component of the network. Known as Reconfigurable Intelligent Surfaces (RIS) or Intelligent Reflecting Surfaces (IRS), this technology allows network operators to dynamically reshape how electromagnetic waves propagate through space.

The Intuition of Passive Beamforming

To understand why an RIS works, it is helpful to contrast it with traditional network infrastructure like active relays or base stations. An active relay receives a signal, amplifies it (along with any background noise), and retransmits it using power-hungry radio frequency (RF) chains containing mixers, power amplifiers, and digital signal processors.

In contrast, as Marco Di Renzo et al. explain, an RIS is a "nearly passive" surface composed of an array of sub-wavelength scattering elements (meta-atoms). Each element is integrated with a simple electronic switch, such as a PIN diode or varactor, controlled by a low-power computing unit like an FPGA. Instead of amplifying the signal, the RIS alters the local boundary conditions of the electromagnetic field. By adjusting the impedance or phase shift of individual elements in real-time, the surface can collectively perform functions such as anomalous reflection, beam focusing, or interference nulling without active RF amplification.

Traditional Reflection (Snell's Law):
Incident Wave \       / Reflected Wave (Angle of Incidence = Angle of Reflection)
               \     /
________________v___v________________

Anomalous RIS Reflection:
Incident Wave \     | Focused/Steered Wave (Phase shifts alter wave front)
               \    |
________________v___v________________
               [RIS Array]

This passive manipulation yields an extraordinary physical advantage: the $N^2$ power scaling law. As demonstrated by Qin Tao et al., an RIS with $N$ reflecting elements can provide a received signal-to-noise ratio (SNR) gain proportional to $N^2$.

The intuition behind this quadratic scaling lies in a two-step physical process:

  1. Aperture Gain: The physical size of the RIS allows it to intercept an amount of electromagnetic energy proportional to the number of elements $N$.
  2. Beamforming Gain: Because the elements are programmed to align the phases of the reflected waves perfectly at the target receiver, the electric fields add up coherently. Since power is proportional to the square of the combined electric field amplitude, this coherent combination introduces another factor of $N$.

By coordinating these reflections to combine constructively at the receiver, as investigated by Chongwen Huang et al., the network can dramatically boost signal strength and eliminate dead zones without increasing the transmitter's power output.

Overcoming the High-Frequency Blockage Problem

The practical urgency for deploying RIS arises from the physics of next-generation wireless bands. To achieve gigabit-and-beyond data rates, 5G-Advanced and 6G networks are moving toward millimeter-wave (mmWave) and Terahertz (THz) frequencies. However, as Christina Chaccour et al. point out, these high-frequency bands suffer from severe atmospheric absorption, high path loss, and an inability to penetrate solid objects. A single concrete wall, a moving vehicle, or even a human body can completely block a THz or mmWave link.

RIS solves this by engineering "virtual line-of-sight" (LoS) paths around physical obstructions. When a direct path between a base station and a user is severed, an RIS mounted on a nearby wall or building can capture the signal and steer it directly to the user. Yuan Xu et al. demonstrate through stochastic geometry that distributing these surfaces across an environment significantly expands coverage and achievable data rates in blockage-heavy scenarios.

This capability is not limited to outdoor cellular networks. For instance, Omid Abari et al. demonstrated how programmable mmWave reflectors can be used to maintain high-bandwidth, low-latency connections for untethered Virtual Reality (VR) headsets, preventing signal drops when a user turns their head or moves their hands in front of the line-of-sight path.

Algorithmic and Structural Innovations

Realizing the promise of programmable environments requires solving a massive optimization problem: how do we configure thousands of passive elements in real-time as users move?

Bypassing Channel Estimation Overhead

Traditional beamforming relies on precise Channel State Information (CSI). However, because an RIS is passive, it cannot transmit or receive pilot signals directly to estimate the channel. Estimating the individual channels from the base station to each RIS element, and from each element to the user, creates an unsustainable signaling overhead.

To circumvent this, Keke Ying et al. proposed a geometric approach that configures the RIS phase shifts using only the macroscopic angles of arrival and departure of the dominant line-of-sight paths, bypassing the need for full CSI. Alternatively, Tao Jiang et al. introduced an active sensing framework using "ping-pong" pilots and deep learning to adaptively learn optimal reflection patterns without explicit channel feedback.

Beyond Diagonal Architectures

Standard RIS designs are "diagonal," meaning each reflecting element operates completely independently of its neighbors. While mathematically simple, this limits the surface's efficiency. Hamad Yahya et al. introduced "Beyond Diagonal RIS" (BD-RIS), an architectural evolution where elements are interconnected via passive reconfigurable networks. This cross-element connection provides additional degrees of freedom, allowing the surface to perform complex wave manipulation like passive Maximum Ratio Transmission (MRT) and highly precise interference nulling in multi-user environments.

Diagonal RIS (Conventional):
[Element 1]   [Element 2]   [Element 3]
    |             |             |
(Isolated phase control for each element)

Beyond Diagonal RIS (BD-RIS):
[Element 1] <---> [Element 2] <---> [Element 3]
     \               /                 /
      <------------->----------------->
(Interconnected elements allow cross-coupling and superior wave shaping)

System-Level Integration

Because an RIS can selectively favor or suppress specific signals, it acts as a powerful force multiplier when paired with other network technologies:

  • Multi-User Access: By combining RIS with Non-Orthogonal Multiple Access (NOMA), Xinyu Gao et al. and Ziwei Liu show that networks can more effectively separate overlapping user signals in the power domain, boosting overall spectral efficiency.
  • Edge Computing and Learning: In mobile edge computing, Paolo Di Lorenzo et al. leverage RIS to optimize the offloading of computational tasks from low-power devices. Similarly, Hang Liu et al. demonstrate that RIS can stabilize and accelerate Over-the-Air Federated Learning by physically balancing the wireless channels of disparate edge devices during model aggregation.
  • Data Freshness: In time-sensitive networks, Yue Xiu demonstrated that deploying an aerial, drone-mounted RIS can minimize the Age of Information (AoI) by dynamically repositioning the reflective link to prioritize stale data streams.

By treating the physical world not as a fixed constraint but as a software-defined variable, reconfigurable intelligent surfaces offer a path toward highly efficient, green wireless infrastructure capable of meeting the extreme spatial and spectral demands of future networks.

History & citationVersion 2 · 2026-10-04

Current · Permanent link

  1. Version 2 (latest) · 2026-10-04 · 236 papers
  2. Version 1 · 2026-10-04 · 260 papers
Cite this direction

Latest version (a living page):

@misc{lacuna_reconfigurable_intelligent_surfaces_6f16dcb5,
  title = {reconfigurable intelligent surfaces},
  author = {{Lacuna}},
  howpublished = {Lacuna research direction},
  year = {2026},
  url = {https://lacuna.tiptreesystems.com/direction/reconfigurable-intelligent-surfaces/rdr_6f16dcb569cc4ce3aaf878e9c4758822},
  note = {Latest version of a living research direction; cite a version for a fixed text. Accessed 2026-11-19},
}

This exact version (version 2, frozen text):

@misc{lacuna_reconfigurable_intelligent_surfaces_6f16dcb5_v2,
  title = {reconfigurable intelligent surfaces},
  author = {{Lacuna}},
  howpublished = {Lacuna research direction},
  year = {2026},
  url = {https://lacuna.tiptreesystems.com/direction/reconfigurable-intelligent-surfaces/rdr_6f16dcb569cc4ce3aaf878e9c4758822/v/2},
  note = {Version 2, frozen 2026-10-04. Accessed 2026-11-19},
}
About

These directions overlap with this one through shared papers.

See something wrong? Report a mistake