Source-linked AI summary
Radio Localization and Mapping with Reconfigurable Intelligent Surfaces
Henk Wymeersch, Jiguang He, Benoît Denis, Antonio Clemente, Markku Juntti
TL;DR
The paper examines RIS-assisted localization and mapping amid uncontrolled physical environments and immature working assumptions. It considers RIS models, control, signaling, and system architecture, arguing that RISs can benefit localization through improved accuracy and extended coverage.
Problem
Uncontrolled physical environments and immature working assumptions remain challenges for RIS-based localization and mapping.
Method
The paper examines RIS models, geometry, control mechanisms, signaling, feedback, and system architecture for localization and mapping.
Results
5G enables location accuracy on the order of 10 m, while Beyond 5G systems can reach around 1 m; one RIS model offers coverage and errors comparable to 2 active BSs.
Takeaways & Limitations
RISs can benefit localization and mapping through improved accuracy or extended coverage while supporting dual communication and localization functions.
Takeaways & Limitations
Progress is hampered by the immaturity of working assumptions, while aspects of the physical environment remain uncontrolled.
Abstract
from arXiv · showhide
5G radio at millimeter wave (mmWave) and beyond 5G concepts at 0.1-1 THz can exploit angle and delay measurements for localization, by the virtue of increased bandwidth and large antenna arrays but are limited in terms of blockage caused by obstacles. Reconfigurable intelligent surfaces (RISs) are seen as a transformative technology that can control the physical propagation environment in which they are embedded by passively reflecting EM waves in preferred directions. Whereas such RISs have been mainly intended for communication purposes, RISs can have great benefits in terms of performance, energy consumption, and cost for localization and mapping. These benefits as well as associated challenges are the main topics of this paper.
INTRODUCTION
5G and beyond-5G radio can improve localization through larger bandwidths, higher frequencies, and antenna arrays, but blockage limits propagation. RISs offer a controllable propagation environment for localization and mapping, motivating broader study of their applications, models, and challenges.
- High-frequency radio propagation is vulnerable to blockage, while RISs can redirect signals through controllable reflection, refraction, or lens-like behavior.
- RIS-based localization and mapping can use multipath constructively, including joint estimation of user and object locations.
- RIS properties and their relationship to environment geometry make them attractive for localization and mapping, an area with limited prior literature.
- The paper takes a broader view by describing technical challenges, a preliminary system vision, and results for applying RISs to localization and mapping.
RADIO LOCALIZATION AND MAPPING
Radio localization and mapping systems are organized around measurements, a reference system, and algorithms. The measurements come from the radio channel between transmitters and receivers.
- A radio localization and mapping system comprises measurements, a reference system, and localization and mapping algorithms.
- Measurements are derived from radio signals between a transmitter and receiver and can typically be obtained from channel estimation.
used for communication. Common location-dependent metrics
Localization uses channel-derived measurements and known references to infer positions, while RIS control can shape signals and potentially improve coverage and accuracy. Performance depends on resolvability, signal properties, and reference-system quality.
- Common location-dependent measurements include received signal strength, ToA, PoA, AoA, and related angular or Doppler quantities.
- Measurement resolution depends on bandwidth and antenna count, whereas accuracy also depends on SNR, waveform properties, and power allocation.
- Reference systems use known anchor states and require considerations such as synchronization and array calibration; their geometry affects accuracy through GDOP.
- Localization traditionally uses LoS measurements, while modern approaches also exploit scattered or reflected NLoS paths and multipath.
- RISs can be integrated into localization systems with known positions and orientations, while users and passive objects may remain unknown or partially known.
- RIS controllers can shape signals and support power allocation and beamforming, but RIS signal design may be less flexible than conventional base-station signals.
- RIS-based SLAM should use base-station signal flexibility and RIS controllability to improve localization and mapping coverage and accuracy.
CHALLENGES AND OPPORTUNITIES
RIS research spans diverse antenna technologies and terminology, while accurate functionality and electromagnetic interaction models remain active challenges. These issues frame opportunities for programmable wireless environments and RIS-based localization and mapping.
- RIS implementations include phased arrays, reflectarrays, transmit arrays, software-defined metasurfaces, and large intelligent surfaces.
- Creating ubiquitous programmable wireless environments requires addressing the differing technologies and terminology used for RISs.
- Accurate models of RIS functionality and electromagnetic-wave interaction remain an active research area.
and phase or phase-only control. Quasi-continuous phase
RIS models must represent geometry, element behavior, hardware impairments, and control resolution across frequency and operating mode. These models create localization and mapping opportunities, but require spatial and temporal consistency.
- Element spacing, impedance matching, scattering properties, oblique incidence, mutual coupling, and losses can affect RIS performance.Transmitarray and reflectarray models should include element scattering properties and relevant electromagnetic effects.
- Hardware impairments include RF losses, limited phase-shifter resolution, active-element performance, and frequency-dependent control constraints.Higher carrier frequencies make hardware impairments more pronounced, affecting required flexibility and control.
- RIS models offer localization and mapping opportunities because their behavior depends on geometry and can differ across RIS technologies.Models used for these tasks should remain spatially and temporally consistent and include relevant passive and active objects.
- RIS models depend on surface location, orientation, extent, frequency band, phase behavior, and operating mode.Models may use quasi-continuous or quantized phase shifts and time delays.
Near-field Propagation
RIS near-field propagation requires modeling wavefront curvature rather than relying only on plane-wave assumptions. That curvature can also provide localization information and support infrastructure-light positioning and SLAM.
- Near-field Propagation: Beyond the Fraunhofer distance, plane-wave assumptions hold, whereas closer propagation produces near-field wavefront curvature.A 20 cm × 20 cm RIS has an approximately 8 meter near-field region.
- Near-field Propagation: Near-field curvature can reduce the need for infrastructure or synchronization in localization.The opportunity arises because the near-field signal provides information beyond conventional far-field angular structure.
- Near-field Propagation: A near-field signal provides information about both angle and position of arrival.This information can be observed by an array that is asynchronous and non-coherent with the transmitter.
- Near-field Propagation: Spherical-wave localization can exploit near-field RIS properties for multipath-aided positioning and simpler SLAM data association.RIS size is identified as one property that can be used in this setting.
requires novel dedicated signal processing methods, as well
RIS-aided localization requires channel estimation and signal processing that connect sparse geometric channel parameters to user position. Receive-mode estimation and limited RIS processing capabilities remain important challenges.
- Localization extracts AoA, AoD, ToA, and path or cluster spreads from the compound channel.These channel parameters are then related to user location through 3D geometry.
- Channel estimation to and from a RIS is challenging when the surface has limited processing capability or few RF chains.This constraint is especially relevant for RIS operation in reflect mode.
- A proposed communication protocol separately estimates LoS and RIS channels by activating different RIS phase patterns while transmitting pilots.The phase-pattern changes introduce delays used in the estimation process.
- Sparse high-frequency channels can be estimated with compressive sensing, while geometric information can serve as partial CSI or CSI statistics.Bayesian methods are needed when user location is only statistically known.
- Localization architectures include uplink, downlink, and sidelink, each with different processing, power, and relative-localization characteristics.Uplink offers richer measurements and BS processing, while downlink can localize multiple users and reduce UE power.
- RIS localization requires calibration and synchronization among position references, together with accurate RIS location and orientation information.Calibration can use over-the-air or wired links, and static RISs can be surveyed once.
step or by the use of GPS signals when available. For mobile
RIS tracking and control must account for signaling, mobility, hardware limits, and energy efficiency. Control updates and the trade-offs among localization architectures remain open design questions.
- RIS-aided localization needs control and feedback signals among network entities and methods for tracking the RIS.Estimated user locations can be fed back to refine RIS settings and selection in later localization steps.
- The trade-offs among uplink, downlink, and sidelink RIS-aided localization remain largely unexplored.Signaling-protocol design is identified as an open opportunity.
- RIS Control: Mobile RISs require dedicated tracking routines, while transmit, receive, and reflect modes impose distinct requirements.Physical RIS placement is also constrained by the environment and legal restrictions.
- RIS Control: RIS control adjusts surface impedances to steer beams, but efficient control depends on latency constraints and hardware properties.Phase-shift accuracy and speed are practically limited, and phase control is often quantized.
- RIS Control: RIS material and hardware properties affect power consumption and overall system energy efficiency.These constraints motivate research on whether control should update at frame or symbol level.
and mapping applications can be supported with low update
RIS control can use prior UE and map information to select surfaces, configure reflections, and support localization and mapping. The resulting designs must account for localization objectives, uncertainty, phase quantization, and waveform flexibility.
- RIS control: A priori UE location and map information can guide RIS activation and phase or amplitude control while directing other RISs away from the UE.Limited UE or BS feedback can support RIS design through predetermined codebooks.
- Design requirements: Localization-oriented control differs from communication-oriented control because localization accuracy, rather than SNR and data rate, is the principal metric.Designs must also be robust to position and orientation errors and uncertainty in the map and UE state.
- RIS control: RISs can optimize GDOP or other localization-relevant metrics and reflect incoming signals toward multiple directions for multi-user support from one base station.Independent reflections can also use different polarizations, frequency bands, or sub-array architectures.
- Mapping workflow: RIS activation schedules can dim or illuminate inaccessible environmental regions, enabling mapping beyond areas directly accessible by the BS.The figure’s flow sends RIS-reflected pilots to UE channel estimation and then SLAM for UE location and local-map determination.
- Design requirements: Finite RIS phase quantization enables low-power, low-complexity control but limits the flexibility of usable codebooks.Dedicated localization waveforms and RIS codebooks should remain flexible enough for accurate angle or delay estimation.
the performance of different codebooks at the UE and RIS,
Codebook and RIS-assisted localization are evaluated through position and orientation errors and through far-field position-error bounds. The paper highlights gains from RIS-aware localization while identifying data association, resolvable paths, mobility, and algorithmic complexity as challenges.
- Far-field comparison: Figure 6 compares far-field position-error bounds across scatterer, reflector, two-BS, and two RIS path-loss-model scenarios.The comparison uses Model 1 with scatter-like per-element loss and Model 2 with reflector-like per-element loss.
- Algorithmic challenges: RIS-aided SLAM must associate detected paths with RISs and passive objects despite clutter and missed detections caused by directional beamforming.The user state includes 3D position, 3D orientation, and clock bias, requiring enough resolvable paths.
- RIS-assisted SLAM: A priori RIS location and orientation reduce data-association hypotheses and allow better localization of passive landmarks and users in monostatic and bistatic configurations.The stated benefits include localization accuracy and service coverage.
- Algorithmic challenges: Bayesian factor-graph and message-passing methods are identified as suitable for RIS-based SLAM because the problem combines heterogeneous signal sources in potentially asymmetric or cooperative settings.The paper describes this as a complex new RIS-based SLAM problem.
Model 2 could even lead to much better performance in terms
The paper argues that RISs can benefit localization and mapping when appropriate physical models and algorithms are developed. It presents an overview of the field, its challenges, research questions, and possible research directions, while emphasizing unresolved feasibility and modeling issues.
- Conclusions and Outlook: RISs can benefit localization and mapping through improved localization performance and physical coverage when appropriate models and algorithms are developed.The paper presents this as a conclusion about RIS-based localization and mapping.
- Conclusions and Outlook: Progress is hampered by immature working assumptions, while integrating and controlling RISs at low cost and using UE location for optimal control remain challenging.These constraints bound the practical readiness of the proposed localization and mapping vision.
- Conclusions and Outlook: Different RIS visions imply distinct physical behaviors, advantages, and drawbacks because technological maturity affects end-to-end power loss over reflected paths.The paper therefore treats RIS performance as dependent on the underlying physical model.
- Conclusions and Outlook: The paper’s overall aim is to provide an up-to-date overview of RIS-based localization and mapping, including its main challenges and prominent research questions.It also outlines potential avenues for answering those questions.
- Conclusions and Outlook: The paper positions RISs as a potential game-changer for next-generation localization and mapping and calls for attention across communication, signal-processing, propagation, and antenna communities.This outlook is framed for 5G or 6G localization.