Reconfigurable Intelligent Surfaces for Rich Scattering Wireless Communications: Recent Experiments, Challenges, and Opportunities

George C. Alexandropoulos, Nir Shlezinger, Philipp del HougneIEEE Communications Magazine 2021 · DOI

TL;DR

This paper explores the exciting potential of Reconfigurable Intelligent Surfaces (RIS) in complex, "rich scattering" wireless environments, moving beyond idealized free-space assumptions. The authors demonstrate how RIS can actively shape these challenging channels to improve communication rates and enable novel localization techniques, paving the way for more robust and intelligent wireless systems.

What Problem Does This Paper Address?

Traditional wireless communication research often assumes relatively "clean" signal paths, where signals travel directly or with a few predictable reflections. However, real-world environments, especially dense urban areas and indoor spaces, are "rich scattering" environments. Here, signals bounce off countless surfaces, creating a complex web of multipath components. This makes it difficult for receivers to distinguish the original signal from its echoes, leading to Inter-Symbol Interference (ISI) and hindering accurate localization. Existing Reconfigurable Intelligent Surfaces (RIS) research primarily focuses on these idealized scenarios, aiming for precise beamforming. This paper addresses the gap by investigating how RIS can be effectively utilized in these inherently complex, highly reverberant environments, where the goal shifts from precise beam steering to manipulating the overall scattering behavior of the environment.

Methods

The paper leverages recent physics-driven experimental explorations of RIS in complex scattering cavities. Two key experimental case studies are presented:

  • Analog Multipath Shaping: An RIS is used to actively shape the Channel Impulse Response (CIR) of a rich scattering environment. This involves configuring the RIS meta-atoms to create constructive interference at specific time instances, effectively tailoring the multipath profile. The RIS configuration is optimized iteratively through trial and error, as analytical modeling is complex.
  • Wave Fingerprinting (WFP) for Localization: The RIS-tunable propagation environment is used as an analog multiplexer. By probing the EM wave field at a single frequency and position with a series of random RIS configurations, a unique "fingerprint" of the environment is generated for each object's location. This fingerprint is then used to train a simple Artificial Neural Network (ANN) for localization. This approach avoids the need for complex optimization of RIS configurations for localization.

Datasets

Not discussed in the paper. The experiments utilize custom-built setups and prototypes.

Key Insights and Ideas

  • RIS as an "Analog Processor": The core idea is that RIS can perform computational tasks by manipulating EM wave propagation, rather than relying solely on traditional electronic processors. This "analog processing" is particularly powerful in rich scattering environments.
  • Shaping the CIR for Capacity: By actively shaping the CIR, RIS can improve the achievable data rates in multipath channels. This is achieved by focusing the signal's energy at desired time instances, effectively mitigating ISI without impacting thermal noise.
  • Wave Fingerprints for Localization: The inherent sensitivity of chaotic wave fields to tiny perturbations can be exploited. Each object's position creates a unique EM wave fingerprint, which can be captured and interpreted by an ANN for localization, even in the absence of line-of-sight.
  • Random Configurations for Localization: Unlike channel shaping, localization via WFP benefits from random RIS configurations, simplifying the experimental setup and reducing the need for complex optimization.
  • ANNs for Complex Environments: Artificial Neural Networks are crucial for interpreting the complex data generated in rich scattering environments, enabling both channel shaping optimization and accurate localization.

Key Results

  • Improved Communication Rates: The RIS-shaped channels demonstrated significant capacity improvements compared to the original channel, with time-coherent focusing of CIR energy leading to notable gains. For instance, specific RIS configurations resulted in impulse-like or bi-modal impulse responses, enhancing data rates.
  • Accurate Localization: The proposed WFP-based localization scheme achieved excellent accuracy, correctly identifying object positions even at very low Signal-to-Noise Ratios (SNRs) (as low as -40 dB) and with a relatively small number of RIS configurations (as few as 100). This highlights the compressive nature of the EM wave propagation in RIS-controlled environments.

Limitations and Future Work

Limitations:

  • The paper acknowledges the complexity of analytically modeling the impact of RIS in rich scattering environments, necessitating iterative optimization for channel shaping.
  • The WFP localization approach, while effective, relies on an assumption of linear combination of WFPs for multiple objects, which holds "to a good approximation" when inter-object scattering is negligible.

Future Work:

  • Exploring the joint optimization of RIS configurations for multiple users in multi-user communication scenarios under rich scattering.
  • Investigating the application of analog multipath shaping on a chip scale.
  • Developing algorithmic designs that jointly optimize RIS-enabled analog processing with transceiver functionalities, while minimizing computational complexity and control signaling overhead.
  • Enhancing localization accuracy by jointly optimizing with conventional schemes or integrating with other communication functionalities.
  • Developing ANNs that can update themselves in-field and transfer knowledge between different localization problems.
  • Leveraging ANNs to capture the impact of RIS on the channel, enabling numerical optimization of iterative procedures.

Summary updated

BibTeX
@article{alexandropoulos2021reconfigurable,
  author = {George C. Alexandropoulos and Nir Shlezinger and Philipp del Hougne},
  title = {{Reconfigurable Intelligent Surfaces for Rich Scattering Wireless Communications: Recent Experiments, Challenges, and Opportunities}},
  journal = {IEEE Communications Magazine},
  year = {2021},
  doi = {10.1109/mcom.001.2001117},
}
Figure