Source-linked AI summary
6G White Paper on Localization and Sensing
Andre Bourdoux, Andre Noll Barreto, Barend van Liempd, Carlos de Lima, Davide Dardari, Didier Belot, Elana-Simona Lohan, Gonzalo Seco-Granados, Hadi Sarieddeen, Henk Wymeersch, Jaakko Suutala, Jani Saloranta, Maxime Guillaud, Minna Isomursu, Mikko Valkama, Muhammad Reza Kahar Aziz, Rafael Berkvens, Tachporn Sanguanpuak, Tommy Svensson, Yang Miao
TL;DR
The white paper addresses gaps between stringent future 6G localization and sensing requirements and existing sensor capabilities. It identifies enabling technologies and key procedural aspects, reporting high spatial and delay resolution and experimental feasibility of 1 deg angular and cm-scale range resolution, with concepts and services positioned to change in the 6G era.
Problem
A significant gap remains between the most stringent requirements for future 6G localization and sensing systems and existing capabilities, motivating more radical mobile sensor technology.
Method
The white paper focuses on localization and sensing procedures by identifying potential enabling technologies and their main features.
Results
High spatial and delay resolution are available through pencil-like beams at mmWave and µmWave bands and ultra-wide bandwidth, while 1 deg angular and cm-scale range resolution is experimentally feasible today.
Takeaways & Limitations
The paper presents concepts, ideas, and services for localization and sensing as central elements that will change in the 6G era.
Abstract
from arXiv · showhide
This white paper explores future localization and sensing opportunities for beyond 5G wireless communication systems by identifying key technology enablers and discussing their underlying challenges, implementation issues, and identifying potential solutions. In addition, we present exciting new opportunities for localization and sensing applications, which will disrupt traditional design principles and revolutionize the way we live, interact with our environment, and do business. Following the trend initiated in the 5G NR systems, 6G will continue to develop towards even higher frequency ranges, wider bandwidths, and massive antenna arrays. In turn, this will enable sensing solutions with very fine range, Doppler and angular resolutions, as well as localization to cm-level degree of accuracy. Moreover, new materials, device types, and reconfigurable surfaces will allow network operators to reshape and control the electromagnetic response of the environment. At the same time, machine learning and artificial intelligence will leverage the unprecedented availability of data and computing resources to tackle the biggest and hardest problems in wireless communication systems. 6G will be truly intelligent wireless systems that will not only provide ubiquitous communication but also empower high accuracy localization and high-resolution sensing services. They will become the catalyst for this revolution by bringing about a unique new set of features and service capabilities, where localization and sensing will coexist with communication, continuously sharing the available resources in time, frequency and space. This white paper concludes by highlighting foundational research challenges, as well as implications and opportunities related to privacy, security, and trust. Addressing these challenges will undoubtedly require an inter-disciplinary and concerted effort from the research community.
I. INTRODUCTION
The white paper frames localization and sensing as central capabilities for 6G, building on higher frequencies, wider bandwidths, massive arrays, intelligent surfaces, beamspace processing, and AI/ML. It surveys enabling technologies, applications, challenges, and research directions for integrating these capabilities into future wireless systems.
- I. INTRODUCTION: The paper focuses on 6G localization and sensing by identifying enabling technologies, assessing environment-aware applications, and recommending trends alongside key research questions.The stated focus covers technology identification, opportunity assessment, and research guidance.
- I. INTRODUCTION: Higher carrier frequencies, wider bandwidths, and larger antenna arrays create opportunities for accurate localization, high-definition imaging, and frequency spectroscopy.The THz range is highlighted for accurate localization, imaging, and spectroscopy.
- I. INTRODUCTION: Localization and sensing support applications including emergency-call localization, through-the-wall detection, navigation, personal radar, and robot and drone tracking.Location information also supports communication-network design, operation, and optimization.
- I. INTRODUCTION: The paper identifies four 6G enablers: new spectrum, intelligent surfaces, intelligent beam-space processing, and AI/ML techniques.These enablers are discussed together with technological challenges and future opportunities.
- I. INTRODUCTION: Higher-frequency systems can provide finer range, Doppler, and angular resolutions, supporting the envisaged localization and sensing applications.The paper links these opportunities to unresolved challenges and open issues that must still be addressed.
- I. INTRODUCTION: 6G radios are expected to use channel bandwidths at least five times larger than 5G, while operation up to 300 GHz is technically considered possible.The paper also discusses propagation effects, hardware integration, and the need for consistent channel models across frequency bands.
B. Intelligent reflective surfaces for enhanced mapping and localization
Intelligent reflective surfaces (IRSs) reshape radio propagation to extend coverage, support NLOS communication, and enhance localization and sensing. Their practical deployment still requires improved channel and material models, hardware implementations, placement strategies, and signal processing.
- Capabilities: IRSs control radio-channel features such as scattering, reflection, and refraction by dynamically adapting surface parameters.The parameters can include phase, amplitude, frequency, and polarization without requiring complex decoding, encoding, or radio-frequency operations.
- Applications: IRSs can facilitate NLOS communication, tracking and surveillance, autonomous localization, and enhanced localization and sensing in mmWave and THz bands.Large intelligent surfaces can also exploit near-field wavefront curvature to improve location accuracy and potentially remove explicit synchronization between reference stations.
- Capabilities: IRS deployments can extend communication range and enhance system performance, especially where high-frequency propagation paths are sparse and losses are large.At high frequencies, IRSs add controlled scattering and can be electronically large while occupying a small physical footprint.
- Applications: Accurate localization and sensing can improve the efficiency of intelligent-surface operation, while passive THz sensing can identify reflecting-surface material types for beamforming decisions.This supports using environmental knowledge across communication, sensing, imaging, and localization applications.
- Challenges: IRS research remains constrained by missing material and propagation models, difficult channel estimation, deployment-location selection, and optimization of joint communication, sensing, and localization.High-frequency channels tend to be low-rank and therefore carry less information; IRSs also lack their own radio resources for transmitting pilot symbols.
C. Beamspace processing for accurate positioning
Beamspace processing uses directional multi-antenna transmission and angular-delay channel information for accurate positioning and sensing in LOS, NLOS, and device-free scenarios. Its effectiveness depends on channel estimation, blockage management, hardware robustness, and balancing accuracy against complexity and latency.
- Processing: Beamspace processing combines directional transmission, multiple spatial streams, and angular-delay channel profiles to support localization and sensing.Collaborative distributed multi-antenna systems can localize passive and active targets over LOS and NLOS paths.
- Localization: LOS angle estimates can directly infer an active user’s location, whereas NLOS multipath estimates locate scatterers and require both AoA and AoD to trace the user.Continuous beamspace measurements can be compared with static references for passive sensing or previous samples for moving-target tracking.
- Target identification: Device-free localization and sensing require environmental indications that distinguish target characteristics from background objects, especially when multiple targets are present.Learning algorithms can identify targets from variations in angular-delay profiles associated with size and dielectric properties.
- Resolution: Narrow mmWave and µmWave beams provide very high spatial resolution, while ultra-wide bandwidth provides very high delay resolution.Combined angular-delay profiles can therefore be used to localize and sense passive targets.
- Challenges: Blockage and deep fading from moving background objects can degrade beamspace signal quality and localization accuracy, motivating coordinated multi-antenna stations and predictive environmental modeling.Potential tools include depth cameras, ray tracing, propagation graphs, and image-based mobility prediction.
- Challenges: Additional open issues include phase noise and non-linearity above 40 GHz, near-field direction estimation, closely spaced target separation, and the trade-off among algorithm complexity, hardware capability, and time consumption.Performance evaluation should include localization accuracy and the Cramér-Rao bound under system specifications.
D. Machine learning for intelligent localization and sensing
Machine learning is positioned as an enabler for localization and sensing when observations are multimodal, indirect, noisy, high-dimensional, or difficult to model analytically. The section emphasizes hybrid and adaptive learning approaches while identifying data, labeling, computation, and non-stationarity as central challenges.
- Motivation: Traditional mathematical models and signal-processing techniques are insufficient alone for many localization problems involving multimodal, indirect, noisy observations and difficult-to-model nonlinear signals.This motivates data-driven methods for modeling system behavior, sensor noise, and uncertainty.
- Capabilities: Machine learning can map high-dimensional low-level measurements, such as massive-MIMO CSI, to higher-level localization and sensing concepts.Predictive models and pattern recognition support feature extraction, target recognition, tracking, mapping, cooperative localization, and autonomous navigation.
- Approaches: Future systems are expected to combine physics-based propagation models, data-driven learning, and sequential Bayesian state-space models.Deep learning and probabilistic methods are presented as routes toward more flexible and accurate localization and sensing approaches.
- Sensing opportunities: Higher carrier frequencies and wider spectral ranges can improve environmental measurement and enable sensing of targets and variables not detectable in currently used bands.AI and ML can extract hidden patterns and integrate weak, noisy signals temporally and spatially.
- Challenges: Established AI and ML techniques are often data-hungry, requiring substantial labeled training data and computing power.Semi-supervised learning, federated learning, and reinforcement learning are identified as possible responses to limited data and adaptive environments.
- Challenges: Localization and sensing environments are increasingly autonomous, non-stationary, and time-evolving, requiring online adaptive machine-learning techniques.Addressing these data and adaptation challenges is essential for highly dynamic, large-scale systems.
III. LOCALIZATION AND SENSING OPPORTUNITIES FOR FUTURE 6G SYSTEMS
The paper envisions 6G localization, sensing, and communication as convergent functions that share resources and mutually support one another. It identifies performance gaps in current standards and proposes broader KPIs and coordinated system design for future applications.
- III. LOCALIZATION AND SENSING OPPORTUNITIES FOR FUTURE 6G SYSTEMS: 6G should integrate localization, sensing, and communication while sharing time-frequency-spatial resources.Coexistence, cooperation, and co-design are identified as mechanisms for this sharing.
- III. LOCALIZATION AND SENSING OPPORTUNITIES FOR FUTURE 6G SYSTEMS: Sensing and location information can guide communication functions such as beamforming and handovers, while communication can support localization and sensing.
- III. LOCALIZATION AND SENSING OPPORTUNITIES FOR FUTURE 6G SYSTEMS: A significant gap remains between stringent use-case requirements and what current 3GPP standards, particularly Release 16, achieve.Release 17 targets may become an objective for future 6G systems.
- III. LOCALIZATION AND SENSING OPPORTUNITIES FOR FUTURE 6G SYSTEMS: Future 6G KPIs should capture UE orientation, environmental maps or features, and power consumption per position fix.The latter is emphasized for positioning in IoT contexts.
- III. LOCALIZATION AND SENSING OPPORTUNITIES FOR FUTURE 6G SYSTEMS: The envisioned technological enablers support new localization and sensing applications within convergent 6G communication systems.Figure 7 illustrates future deployment scenarios and related opportunities.
A. THz imaging
THz imaging offers fine resolution and compact sensing opportunities across radar, material sensing, biomedical monitoring, and imaging. The section contrasts mature but costly bolometer systems with lower-cost CMOS alternatives and outlines remaining integration challenges.
- A. THz imaging: THz and mmWave sensing offer very fine resolution in range, angle, and Doppler dimensions.
- Passive imaging: MOS-based THz image sensors have 1000 times lower current sensitivity than bolometer-based sensors but remain viable cost-effective alternatives.
- Active imaging: THz sensing can identify material properties through spectral fingerprints, spectroscopy, and carrier selection near resonant frequencies.Carrier-based sensing can be integrated with THz communications, although joint operation requires further signal-processing advances.
- Biomedical applications: THz imaging could support compact, contactless biomedical monitoring, including continuous physiological and biochemical measurements.Current biomedical THz systems are costly and use bulky optical components.
B. Simultaneous localization and mapping
6G radio-based SLAM aims to estimate sensor trajectories and landmark states while enabling automatic indoor-map construction from pervasive mobile devices. Its main obstacles are radio-specific measurement uncertainty, complex state models, and computational demands.
- B. Simultaneous localization and mapping: SLAM jointly recovers sensor states, trajectories, and landmark states from measurements collected as a sensor moves.
- Challenges: Key implementation challenges include low-complexity algorithms for mobile devices, hardware impairments, and high computational complexity from unknown data associations.Distributed mobile edge computing is suggested as possible support.
- B. Simultaneous localization and mapping: Radio-based 6G SLAM is harder than conventional laser- or camera-based SLAM because measurements are less accurate and affected by sidelobes and multipath.Ad hoc measurement models tailored to radio propagation are therefore needed.
- B. Simultaneous localization and mapping: 6G SLAM must model sensor position and 3D orientation alongside increasingly fine-grained environmental landmarks.
- B. Simultaneous localization and mapping: THz imaging can reduce the gap between LIDAR-based and radio-based SLAM by providing geometry side information for accurate indoor maps.Pervasive devices such as smartphones, smart glasses, and XR devices could contribute radio measurements.
- B. Simultaneous localization and mapping: Automatically updated indoor maps could enable infrastructure-less and map-less localization, AR/VR/MR applications, and autonomous vehicle and drone localization.
C. Passive sensing using transmitters of opportunity
Passive sensing reuses uncontrolled wireless transmissions to detect and characterize objects without cooperation from targets. It enables applications from human monitoring and localization to imaging, but generally sacrifices performance and resolution relative to active radar.
- C. Passive sensing using transmitters of opportunity: Passive sensing uses signals from transmitters of opportunity in a bi-static radar arrangement without controlling or synchronizing the transmitter.Receivers use prior knowledge of the wireless standard and frame structure to extract environmental information.
- C. Passive sensing using transmitters of opportunity: Passive sensing supports vital-sign, fall, presence, intruder, activity-recognition, localization, environmental-mapping, and through-the-wall applications.
- C. Passive sensing using transmitters of opportunity: Examples include through-the-wall imaging with Wi-Fi, 2D passive radar imaging using digital video broadcasting signals, and road-user localization with 5G signals.
- C. Passive sensing using transmitters of opportunity: Passive sensing seeks range-Doppler-angle radar data cubes, but typically does not achieve the quality and SNR of conventional active radar.
D. Active sensing with radar and communications convergence
6G is expected to converge radar and communications through shared waveforms, spectrum, hardware, and processing. This integration offers efficiency and flexibility but requires resolving performance trade-offs, interference, hardware, and resource-management challenges.
- Challenges: Dynamic automotive and consumer scenarios make simple coexistence less suitable, increasing the need for cooperation, interference management, and adaptive radio-resource allocation.Communication can support radar through sensor information exchange, while radar can aid beam tracking by detecting obstacles and reflectors.
- Convergence strategies: 6G is likely to use co-designed waveforms for radar and communications rather than operating the systems independently.Co-design employs a single waveform for both functions and is identified as the likely path for 6G systems.
- Convergence strategies: Shared radar-communications systems can improve spectrum flexibility, reduce hardware and integration costs, and lower energy consumption.These benefits arise from reusing hardware and integrating both functions in one system.
- Challenges: Active radar can provide more precise results using the transmitted signal as a reference, but strong receiver self-interference creates hardware and signal-processing challenges.In-band full-duplexing and coordinated antenna, waveform, and processing design are needed to reduce interference.
- Design requirements: Radar and communications convergence targets radar range, velocity, azimuth, and elevation performance alongside communication spectral and power efficiency.The waveform must remain flexible across sensing and communication needs, including short- versus long-range operation and different resolutions.
- Challenges: A shared waveform must balance conflicting radar and communications requirements in environments with high mobility, multiple targets, and clutter.Improved radar detection and estimation algorithms are needed for these wireless communication scenarios.
E. Channel charting
Channel charting applies unsupervised dimensionality reduction to CSI to create a consistent pseudo-location map. It supports network functions without requiring prior environmental geometry or actual user-position estimation, but its assumptions create open challenges at higher frequencies.
- Approach: Channel charting applies unsupervised dimensionality reduction to CSI to create a virtual chart on which users can be located and tracked.The resulting pseudo-location is consistent across users and over time.
- Applications: Pseudo-location tracking can support predictive RRM, rate adaptation, handover, mmWave beam association and tracking, and UE grouping.These functions use the charted relationships among users and channel conditions.
- Challenges: Channel charting is not a universal replacement for true position and may require joint processing across multiple transmission points at higher frequencies dominated by line-of-sight propagation.The approach was initially developed for massive MIMO settings with rich scattering and stable propagation features.
- Benefits: Channel charting can self-configure without prior area maps or surrounding-building geometry, which is useful for temporary or emergency networks.Its unsupervised operation avoids requiring prior information about the scattering environment.
- Benefits: Pseudo-location can support privacy-preserving applications such as contact tracing without estimating the user’s actual position.The chart provides a relative, consistent representation rather than a directly mapped geographic location.
- Challenges: Open machine-learning challenges include lifelong learning and designing features suited to samples acquired in a given environment.These questions follow from channel charting’s machine-learning foundations.
F. Context-aware localization systems
Context-aware localization systems use contextual history, distributed storage and processing, and multimodal connectivity to adapt services and localization methods. Their deployment depends on standardized context parameters, security solutions, and governance across heterogeneous data sources.
- Applications: Context-aware applications can communicate more efficiently by predicting data transmission and recognizing when communication is ideal.The paper associates this with lower energy use and better throughput.
- Context awareness: Distributed storage and processing can move personalization and sensor-data computation to network locations where resources are fast and feasible.Sensor-data fusion can detect trends and deviations, including in healthcare scenarios.
- Multimodal localization: Context-aware multimodal localization lets devices switch communication technologies according to current context, reducing power consumption and improving quality of service.Devices can also select local or private technologies instead of national or public technologies.
- Context awareness: 6G context-awareness can detect trends and deviations by storing and processing temporal context data.This supports applications such as hyper-personalization and real-time detection of changes in a person’s status.
- Challenges: Advanced temporal context detection combines highly personal physiological data with open public data, creating security concerns across distributed infrastructure.The paper calls for intelligent security solutions for storing and processing these data.
- Challenges: Distributed context-awareness requires standardized context parameters and interpretation rules across heterogeneous data sources.Without them, connectivity infrastructures cannot translate context into application assets.
G. Security, privacy and trust for localization systems
The expansion of network- and cloud-based localization increases the number of stakeholders and raises trust, privacy, and security concerns. Proposed opportunities include secure location information for digital interactions, but emerging sensing and fingerprinting capabilities can expose sensitive location and identity information.
- Trust and governance: Wider network- and cloud-based localization will involve more stakeholders and increase trust requirements for network, cloud, and location-based service providers.This expands the positioning chain beyond device- and edge-based solutions.
- Opportunities: New security, privacy, and trust solutions could make secure location information usable as a security parameter for digital interactions.Examples include automated driving, health monitoring, social media, and surveillance systems.
- Opportunities: Network- and cloud-based localization may help reduce user-device power consumption and support massive location-based IoT connectivity.The paper presents these as motivations for moving localization functions into network and cloud infrastructure.
- Privacy and security risks: Provider vulnerabilities could expose users’ location patterns to misuse including identity theft, burglaries, toll avoidance, and stalking.The risks arise from attackers extracting location information from service providers.
- Privacy and security risks: THz remote sensing and see-through imaging on handheld devices and wearables could be privacy-invasive.The paper identifies these concerns in connection with future THz communications.
- Privacy and security risks: High-resolution radio sensing combined with machine learning could identify devices through radio-frequency fingerprinting even when they are not transmitting.This creates an additional privacy concern beyond direct location disclosure.
IV. SUMMARY AND RESEARCH QUESTIONS
The white paper identifies technological enablers and emerging applications for 6G localization and sensing, then frames foundational research questions for achieving accurate, efficient, and convergent systems.
- 6G localization and sensing are enabled by new high-frequency spectrum, intelligent reflective surfaces, beam-space processing, AI, and advanced signal processing.These enablers address spectrum expansion, environmental control, tracking and mapping, data-driven wireless intelligence, and convergent communication-radar applications.
- The paper highlights biomedical and security imaging, indoor mapping, passive sensing, location-informed big-data services, and sensing-localization-communication cooperation as emerging opportunities.It also identifies location information as a means to boost security and trust in 6G connectivity solutions.
- Research questions: A central research question is how 6G enablers can achieve cm-level positioning and high-resolution 3D sensing or imaging.The paper presents this as the first of several foundational questions for future localization and sensing systems.
- Research questions: Other questions concern waveform designs that let communication, localization, and sensing share time, frequency, and space resources efficiently.The paper treats convergence across these functions as a core design challenge.
- Research questions: Further research must address energy-efficient high-accuracy operation at high frequencies with mobility, including real-time energy-efficient AI/ML methods.The paper links these questions to the availability of unprecedented data and computing resources.
- Research questions: The paper also asks how to bridge the quality and accuracy gap between passive and active sensing.