PaperlayerView exact revision on arXiv

Source-linked AI summary

An Overview on IEEE 802.11bf: WLAN Sensing

Rui Du, Haocheng Hua, Hailiang Xie, Xianxin Song, Zhonghao Lyu, Mengshi Hu, Narengerile, Yan Xin, Stephen McCann, Michael Montemurro, Tony Xiao Han, Jie Xu

arXiv:2310.17661v1eess.SPcs.NI

TL;DR

Communication-focused WLAN standards may not meet emerging sensing requirements, despite Wi-Fi’s use for detection, localization, and recognition. This overview synthesizes IEEE 802.11bf’s standardization, use cases, sensing procedures, candidate features, evaluation methods, and research directions. It describes a sensing-oriented amendment intended to improve WLAN sensing while minimizing effects on communications.

Abstract

from arXiv · show

With recent advancements, the wireless local area network (WLAN) or wireless fidelity (Wi-Fi) technology has been successfully utilized to realize sensing functionalities such as detection, localization, and recognition. However, the WLANs standards are developed mainly for the purpose of communication, and thus may not be able to meet the stringent requirements for emerging sensing applications. To resolve this issue, a new Task Group (TG), namely IEEE 802.11bf, has been established by the IEEE 802.11 working group, with the objective of creating a new amendment to the WLAN standard to meet advanced sensing requirements while minimizing the effect on communications. This paper provides a comprehensive overview on the up-to-date efforts in the IEEE 802.11bf TG. First, we introduce the definition of the 802.11bf amendment and its formation and standardization timeline. Next, we discuss the WLAN sensing use cases with the corresponding key performance indicator (KPI) requirements. After reviewing previous WLAN sensing research based on communication-oriented WLAN standards, we identify their limitations and underscore the practical need for the new sensing-oriented amendment in 802.11bf. Furthermore, we discuss the WLAN sensing framework and procedure used for measurement acquisition, by considering both sensing at sub-7GHz and directional multi-gigabit (DMG) sensing at 60 GHz, respectively, and address their shared features, similarities, and differences. In addition, we present various candidate technical features for IEEE 802.11bf, including waveform/sequence design, feedback types, as well as quantization and compression techniques. We also describe the methodologies and the channel modeling used by the IEEE 802.11bf TG for evaluation. Finally, we discuss the challenges and future research directions to motivate more research endeavors towards this field in details.

I. INTRODUCTION

WLAN sensing reuses Wi-Fi signals and infrastructure to infer properties of surrounding environments, supporting applications from presence detection to tracking. The paper motivates IEEE 802.11bf as a sensing-oriented amendment to address limitations of communication-focused WLAN standards, then surveys its scope and technical agenda.

Motivation

  • WLAN sensing processes received Wi-Fi signals to detect obstructions and interpret target movement, supporting applications such as gesture control, fall detection, tracking, imaging, activity recognition, and vital-sign monitoring.The paper notes commercial motion-detection products and identifies residential, enterprise, indoor, and outdoor settings.Source: §I · p. 1

Prior WLAN sensing

  • CSI provides channel amplitude and phase across subcarriers, enabling extraction of multipath parameters such as angle of arrival, time of flight, and Doppler shift.RSSI-based methods are easier and lower-cost to implement, but multipath and imperfect AGC can degrade their performance.Source: §I · pp. 1–2

WLAN sensing context

  • Compared with several alternative sensing technologies, WLAN can reuse widely available Wi-Fi infrastructure and avoid extra sensing devices, while 6G cellular and WLAN sensing are described as complementary across outdoor and indoor settings.The paper contrasts WLAN’s indoor utility with cellular networks’ outdoor advantages.Source: §I · p. 2

Limitations and need for 802.11bf

  • COTS WLAN sensing is hindered by unavailable user-side CSI, unknown transmission adaptations, limited single-node sensing, and communication-oriented metrics and compensation.The paper calls for unified channel-measurement interfaces, transceiver negotiation, multi-device coordination, sensing metrics, and enhanced sounding and transmission protocols.Source: §I · p. 3

Need for 802.11bf

  • IEEE 802.11bf specifies sensing-oriented MAC and PHY modifications to improve sensing while limiting effects on communications; sub-7GHz changes are MAC-layer, while 60 GHz changes include PHY and MAC layers.The 60 GHz provisions modify DMG and EDMG implementations.Source: §I · p. 3

Paper scope

  • The overview covers formation and standardization, use cases and KPI requirements, sub-7GHz and DMG measurement procedures, candidate technical features, evaluation methods, channel models, and future directions.It positions these details as extending prior overviews that omitted combinations of DMG procedures, sub-7GHz sensing, technical features, or broad future directions.Source: §I · pp. 4–5

II. IEEE 802.11BF: FORMATION, DEFINITION, AND TIMELINE

The section traces IEEE 802.11bf from its standards-group formation to its sensing definition and core amendment functions. It also identifies the intended operating bands and compatibility with legacy devices.

A. Formation

  • WLAN sensing was discussed in the IEEE 802.11 WNG SC in July 2019, followed by a TIG in October and approval of a Study Group in November 2019.The SG developed the Project Authorization Request and Criteria for Standards Development documents.Source: A. Formation · pp. 5–6
  • IEEE 802.11bf was officially established in September 2020 after approval of its PAR and CSD, enabling concrete work toward a WLAN sensing amendment.The PAR defines a proposed project’s motivation, scope, and content; the CSD details sponsor requirements agreed with the working group.Source: A. Formation · p. 6

B. Definition

  • WLAN sensing uses signals received from sensing-capable STAs to determine target features such as range, velocity, angle, motion, presence, proximity, or gesture.The stated targets include objects, people, and animals in environments such as rooms, houses, vehicles, and enterprises.Source: B. Definition · p. 6
  • The amendment specifies MAC modifications and DMG/EDMG PHY modifications for license-exempt bands from 1 to 7.125 GHz and above 45 GHz.It is intended to support backward compatibility and coexistence with existing or legacy IEEE 802.11 devices operating in the same band.Source: B. Definition · p. 6
  • STAs are to announce sensing capabilities, request and set up measurement transmissions, identify usable transmissions, and exchange sensing feedback and information.Sensing measurements may use requested transmissions, unsolicited transmissions, or both.Source: B. Definition · p. 6
  • The amendment provides a MAC service interface through which higher layers can request and retrieve WLAN sensing measurements.An STA has IEEE 802.11 MAC and PHY interfaces to the wireless medium and may be an AP, laptop, or WLAN-enabled phone.Source: B. Definition · p. 6

C. Timeline

The timeline passage describes IEEE 802.11bf’s draft and anticipated approval and deployment milestones. The use-case material introduces sensing-oriented KPIs, including measurement quality, coverage, timing, and detection performance.

C. Timeline

  • Draft 0.1 was released in April 2022; submission for an IEEE SA ballot was expected as early as March 2024, publication in March 2025, and deployment by the end of 2025.The latter milestones are stated as expectations in the paper.Source: C. Timeline · pp. 6–7
  • During the TIG/SG phase from September 2019 to September 2020, proposals on channel models, sequence design, and measurement procedures were presented and shown technically feasible.Technical ballots were described as the subsequent process for voting on the draft and attaching revision comments.Source: C. Timeline · p. 6

A. Sensing-oriented KPIs

  • IEEE 802.11bf uses sensing-oriented KPIs rather than relying only on communication metrics, with accuracy defined as the difference between estimated range, angle, or velocity and ground truth.The listed KPIs also include latency, refresh rate, simultaneous targets, range coverage, field of view, and range, angular, and velocity resolution.Source: A. Sensing-oriented KPIs · p. 7
  • Range coverage is the maximum target distance at which received SNR exceeds a threshold, conventionally 10 dB or 13 dB.Field of view describes angular sensing coverage, while resolution measures the minimum separation between distinguishable targets.Source: A. Sensing-oriented KPIs · p. 7
  • Probability of detection is the ratio of correct predictions to all possible predictions, for tasks including gesture and presence detection and body-activity detection such as breathing.The listed prediction tasks also include real-person detection.Source: A. Sensing-oriented KPIs · p. 7

B. Key Use Cases and KPI Requirements

The section relates WLAN sensing use cases to performance requirements, ranging from presence and activity recognition to localization, healthcare, and 3D perception. It describes distinct sensing challenges and identifies stringent accuracy needs for some applications.

1) Presence Detection:

  • Presence detection must distinguish stationary people from empty-room conditions, where CSI can appear similar; its maximum range coverage requirement is 10–15 m, with multi-person resolution requirements also specified.For meeting-room presence detection and counting, the range coverage requirement can generally be set as 10 m.Source: 1) Presence Detection · p. 7

2) Activities Recognition:

  • Human activity recognition uses CSI-derived patterns such as Doppler spectrum, target speed, and amplitude, but activities vary by location and often continue without clear pauses for segmentation.The paper associates sub-7GHz with large-motion recognition and 60 GHz with small-motion recognition for higher resolution and improved recognition accuracy.Source: 2) Activities Recognition · pp. 7–8

3) Human Target Localization and Tracking:

  • Human localization estimates target position and tracking follows changes in position over time; the stated range-accuracy KPI is within 0.2 m, for which 60 GHz sensing is needed.Fingerprinting is simple to deploy but its database construction is time-consuming, while geometric methods generalize better but remain limited by antenna count and bandwidth.Source: 3) Human Target Localization and… · p. 8

4) Healthcare:

  • Healthcare sensing estimates respiration and heartbeat from CSI vibration patterns, while sneeze sensing uses Doppler analysis of droplets; velocity accuracy of at least 0.1 m/s is generally needed.Estimating respiration for multiple targets remains challenging.Source: 4) Healthcare · p. 8

5) 3D Vision:

  • WLAN-based 3D vision includes environment mapping and human-skeleton perception, with the paper identifying 60 GHz as the applicable band and 0.01 m range, 0.1 m/s velocity, and 2-degree angular accuracy requirements.These are described as the most demanding requirements among the use cases.Source: 5) 3D Vision · p. 8

C. Motivation of WLAN Sensing Amendment

Communication-oriented WLAN standards leave sensing with unstable measurements, inaccessible or non-unified CSI, and limited coordination. IEEE 802.11bf proposes standardized operations and device roles, including sub-7GHz and 60 GHz procedures, to address these gaps.

C. Motivation of WLAN Sensing Amendment

  • Communication-oriented frames and pilots, inaccessible CSI, and ad hoc acquisition and coordination leave current WLAN sensing short of use-case KPIs.The amendment aims to standardize sensing-specific operations and support coordinated sensing while maintaining or improving communication performance.Source: C. Motivation of WLAN Sensing Amendment · p. 8
  • The amendment targets stable CSI acquisition, inter-STA cooperation, an algorithm-validation platform, and high-band DMG sensing for demanding resolution and accuracy.The overview also identifies improved sensing and communication performance as a benefit of greater STA cooperation.Source: C. Motivation of WLAN Sensing Amendment · pp. 8–9

IV. IEEE 802.11BF SENSING FRAMEWORK AND PROCEDURE

  • Both sub-7GHz and 60 GHz procedures share capabilities exchange, a measurement session, measurement exchange, and session termination, but their exchange variants differ.Sub-7GHz has trigger-based and non-trigger-based variants; DMG sensing has several types, including monostatic, bistatic, multistatic, and passive.Source: §IV · p. 9

A. Transceiver Roles and Specific Definitions

  • Initiator/responder roles describe who starts or participates in sensing, while TX/RX roles describe who transmits or receives sensing PPDUs; a STA may hold multiple roles.An initiator may be TX, RX, both, or neither, with measurements sometimes obtained through responder feedback.Source: A. Transceiver Roles and Specific… · p. 10

B. Sub-7GHz Sensing Procedure

Sub-7GHz sensing uses a four-stage procedure that negotiates capabilities and session attributes before measurements and termination. Its trigger-based exchange polls responders, performs role-dependent sounding, and supports several reporting options.

B. Sub-7GHz Sensing Procedure

  • The sub-7GHz procedure comprises capabilities exchange, measurement session, measurement exchange, and session termination.Sessions establish operational attributes; measurement and session IDs distinguish exchanges and attribute sets.Source: B. Sub-7GHz Sensing Procedure · pp. 10–11

1) TB Sensing Measurement Exchange:

  • In trigger-based exchanges, an AP initiator polls responder STAs before optional downlink or uplink sounding, with the sounding direction determined by TX/RX roles.Polling checks responder availability; the NDP announcement and trigger-frame sounding phases are included when applicable responders answer.Source: 1) TB Sensing Measurement Exchange · p. 12
  • Sensing reports may be immediate or delayed; delayed feedback can combine reports from multiple sensing sessions.Immediate reports concern the current exchange, whereas delayed reports can concern the previous exchange.Source: 1) TB Sensing Measurement Exchange · p. 12
  • Optional threshold-based reporting sends CSI-variation feedback first, then requests measurement reports only from responders whose variation meets their assigned thresholds.Thresholds are set by the initiator and may differ among responders.Source: 1) TB Sensing Measurement Exchange · pp. 12–13, Figure 10 · p. 13

2) Non-TB Sensing Measurement Exchange:

Non-trigger-based sensing lets a non-AP STA initiate an exchange with an AP responder. DMG sensing adapts WLAN procedures for directional 60 GHz operation, with monostatic sensing placing transmission and reception at one STA.

2) Non-TB Sensing Measurement Exchange:

  • In non-trigger-based exchanges, a non-AP STA initiates and an AP responds; the initiator configures the NDP exchange, with sounding and feedback depending on its TX/RX role.The responder sends a corresponding NDP, and the opposite-direction NDP may be kept at minimum length when it is not used for sensing detection.Source: 2) Non-TB Sensing Measurement Exchange · pp. 13–14

C. DMG Sensing Procedure

  • DMG sensing extends WLAN sensing to directional 60 GHz operation, where wider bandwidth and shorter wavelengths enable higher range and angular resolution.Its procedure uses capabilities exchange, a measurement session, sensing bursts and exchanges, and session termination; prior beamforming training is assumed.Source: C. DMG Sensing Procedure · p. 14

1) Monostatic Sensing:

  • In monostatic sensing, one (E)DMG STA is both sensing TX and sensing RX, so uninvolved STAs need not interoperate in the sensing exchange.An uninvolved STA may measure the PPDU at the PHY layer, but discards it at the MAC layer if it is not addressed to that STA.Source: 1) Monostatic Sensing · p. 14

2) Bistatic Sensing:

DMG sensing ranges from a two-STA bistatic exchange to coordinated multi-device configurations and passive sensing. The procedures address beam training, synchronization, scheduling, and the information needed to interpret received signals.

2) Bistatic Sensing:

  • Bistatic sensing uses distinct TX and RX STAs; BRP frames with TRN fields support channel measurement and beam sweeping, with reporting dependent on which role initiates.A receive-role initiator has no reporting phase in this exchange.Source: 2) Bistatic Sensing · pp. 14–15

3) Multistatic Sensing:

  • Multistatic sensing uses multiple RXs, requiring synchronization and an initiation phase to prepare responders; the initiator later polls them for reports in a predefined order.The multi-static PPDU carries synchronization fields for the different responders.Source: 3) Multistatic Sensing · p. 15

4) Monostatic Sensing with Coordination:

  • Coordinated monostatic sensing schedules multiple monostatic devices, but each STA uses its own clock, avoiding inter-STA synchronization overhead.Sounding for responders may be sequential or simultaneous.Source: 4) Monostatic Sensing with Coordination · p. 15

5) Bistatic Sensing with Coordination:

6) Passive Sensing:

  • Passive sensing reuses non-sensing transmissions: downlink methods use beacon frames, while an uplink method uses SSW or Short SSW frames.STAs request beacon directions and AP location for downlink interpretation; an AP can request sector directions and STA location for uplink sensing.Source: 6) Passive Sensing · pp. 16–17

D. Sensing by Proxy Procedure

The section describes how sensing by proxy and a presence-detection example work, then compares sub-7GHz and DMG sensing challenges, performance characteristics, and use cases.

D. Sensing by Proxy Procedure

  • An SBP initiator requests an AP to sense on its behalf; the AP performs sensing with non-AP STAs and reports measurements back, while the initiator may also participate as a responder.This procedure can let a non-AP STA obtain measurements for detecting and tracking environmental changes.Source: D. Sensing by Proxy Procedure · p. 17

E. Presence Detection under IEEE 802.11bf sub-7GHz Sensing

F. Similarities and Differences between Sub-7GHz Sensing and DMG Sensing

  • Both bands face CSI estimation errors and stability challenges; DMG uses full-duplex operation to mitigate synchronization issues, while sub-7GHz relies on algorithm design, and stable transmit power plus RX-gain reporting are suggested for robustness.The estimated CSI includes transmit- and receive-RF effects as well as physical CSI.Source: F. Similarities and Differences… · pp. 17–18
  • Sub-7GHz targets larger coverage, larger targets, and significant movement, whereas DMG is suited to smaller coverage, smaller targets, and fine movement; their bandwidths and antenna configurations also differ.Sub-7GHz supports up to 320 MHz continuous bandwidth and up to 8 antennas/streams; DMG has 1.76 GHz per channel and typically uses directional antennas.Source: F. Similarities and Differences… · p. 18

V. IEEE 802.11BF: CANDIDATE TECHNICAL FEATURES

The section surveys candidate WLAN sensing features, focusing here on waveform and sequence design. It describes ambiguity-based evaluation and synchronization sequences designed for multi-STA sensing.

V. IEEE 802.11BF: CANDIDATE TECHNICAL FEATURES

  • Communication-oriented preambles prioritize communication properties such as PAPR rather than sensing properties such as angle and range resolution or Doppler frequency, motivating new waveform and sequence designs and evaluation metrics.The candidate-features discussion focuses on technologies that attracted attention in IEEE 802.11bf meetings.Source: A. Waveform and Sequence Design · p. 18, §V · p. 18

A. Waveform and Sequence Design

  • Ambiguity functions evaluate waveforms through matched filtering and produce a range-Doppler map; an autocorrelation function uses the same sequence, while a cross-ambiguity function uses different sequences.The sequences are evaluated over time delay and Doppler frequency.Source: A. Waveform and Sequence Design · p. 18
  • LAZ/ZAZ evaluates a local ambiguity function bounded by maximum Doppler frequency and time delay; the example reports relatively low autocorrelation sidelobes and cross-ambiguity magnitudes in the region of interest.For DMG sensing, auto/cross correlation serves as a surrogate for local ambiguity-function evaluation, and IEEE 802.11bf currently uses it to evaluate waveform designs.Source: A. Waveform and Sequence Design · p. 19
  • A proposed synchronization-sequence structure for multistatic sensing assigns different rows of the Sync subfields to different STAs.The row index denotes the STA.Source: A. Waveform and Sequence Design · p. 19, Figure 18 · p. 19
  • The proposed synchronization-sequence design supports sensing by a maximum of 8 different STAs, and the sequences have zero cross-correlation within each group of four and low cross-correlation between the groups.The reported comparison concerns the maximum sidelobe of same-sequence autocorrelation and the maximum cross-correlation magnitude of different sequences.Source: Figure 19 · p. 19, A. Waveform and Sequence Design · p. 20

B. Feedback Types

The section reviews candidate sensing-feedback representations for sub-7GHz and 60 GHz, then compares quantization approaches that balance feedback overhead, computational complexity, and accuracy.

1) Sub-7GHz Sensing:

  • Sub-7GHz feedback candidates include full or partial CSI, TPDP/TCIR, and frequency-domain differential quantization, with differential reporting reducing overhead while retaining measurement integrity.Full CSI preserves environmental information; partial CSI can report only amplitude or phase, while TPDP/TCIR reports early CIR values suited to sensing ranges of a few dozen meters.Source: 1) Sub-7GHz Sensing · p. 20, Figure 20 · p. 21

2) DMG Sensing:

  • DMG sensing can report CIR measurements, a 2D–4D range-Doppler-angular map, or target parameters such as position and Doppler.Target parameters avoid the high feedback overhead of directly reporting R-D-A maps.Source: 2) DMG Sensing · p. 20, Figure 21 · p. 21

C. Quantization and Compression

  • IEEE 802.11n CSI quantization can be reused, but repeated linear-to-dB conversions and multiplication or division make its procedure computationally demanding.Quantization reduces feedback overhead but introduces quantization error.Source: C. Quantization and Compression · pp. 20–21
  • Uniform linear and power-of-two scaling simplify quantization, but simplified and power-of-two scaling increase relative quantization error by reducing the quantization dynamic range.Power-of-two scaling permits binary shifts in hardware or software.Source: C. Quantization and Compression · p. 21
  • Fractional scaling chooses α from predefined values and can constrain β to powers of two; expanding the α set can approach IEEE 802.11n real-value scaling performance.The method seeks a scaling factor that maximizes α/β subject to a stated inequality.Source: C. Quantization and Compression · pp. 21–22
  • The candidate technical features discussed remain unfrozen, leaving subsequent research and proposals able to affect the amendment’s finalization.The reviewed candidates include feedback types and quantization and compression methods.Source: C. Quantization and Compression · p. 22

VI. SIMULATION AND CHANNEL MODEL

The evaluation methodology defines controlled indoor simulation scenarios and evaluates sensing PHY performance through estimated-parameter accuracy, using RMSE-based comparisons.

VI. SIMULATION AND CHANNEL MODEL

  • IEEE 802.11bf documents an evaluation methodology and a channel-model document to support proposal evaluation and development of the draft amendment.The methodology summarizes PHY performance, a limited set of simulation scenarios, and parameters for evaluating contributions.Source: §VI · p. 22, A. Evaluation Methodology · p. 22

A. Evaluation Methodology

  • The proposed simulation scenarios are dense-multipath indoor living rooms and conference rooms with one TX-RX pair and a moving target.Directional or isotropic antennas may be used, and SISO communication is assumed.Source: A. Evaluation Methodology · p. 22
  • Estimated-parameter accuracy, measured by RMSE against ground truth, replaces throughput and packet error rate as the sensing PHY performance metric.The methodology compares accuracy versus SNR curves and accuracy histograms across operation modes and scenarios.Source: A. Evaluation Methodology · p. 22

B. Channel Model

The section compares ray-tracing and data-driven hybrid channel models for WLAN sensing and describes a link-level platform for evaluating sensing performance.

B. Channel Model

  • The deterministic ray-tracing model generates time-variant rays and CIRs from scenario, target, and antenna configurations, but has high computational complexity and cannot represent real-measurement sensing uncertainty.The DDHC hybrid separates target-related rays generated by ray tracing from target-unrelated rays generated with an autoregressive model, with real datasets refining parameters through KL-divergence minimization.Source: B. Channel Model · pp. 22–23, Figure 22 · p. 23, Figure 23 · p. 23

C. Link-Level Simulation

  • The WLAN sensing link-level simulation combines waveform generation, transmitter, environment/channel, receiver, and signal-processing and performance-evaluation modules.It estimates target parameters against simulation ground truth to calculate sensing accuracy, enabling joint validation of waveform, transceiver, channel, and sensing algorithms.Source: C. Link-Level Simulation · pp. 23–24

B. Channel Model

  • DDHC-based and real-world spectrograms show motion peaks, unlike smoother ray-tracing spectrograms, and DDHC sensing performance is close to real-world data and significantly outperforms ray tracing.The comparison uses a conference-room monostatic setup with a person walking, at both sub-7GHz and 60 GHz.Source: B. Channel Model · p. 23, Figure 24 · p. 24

C. Link-Level Simulation

  • Evaluation uses RMSE-based accuracy measures, including accuracy-versus-SNR curves and accuracy histograms for scenarios such as static and moving target estimation.These criteria characterize sensing PHY performance.Source: C. Link-Level Simulation · p. 24

VII. FUTURE RESEARCH DIRECTIONS

The future directions address sensing across coordinated links and frequencies, AI-assisted sensing, and edge learning and inference. The paper identifies outstanding design questions in feedback, processing, resource management, and adaptation to real Wi-Fi environments.

A. Multi-Link Sensing

  • Multi-link devices could improve sensing and support new applications, including fast-moving-target sensing through separate sensing and reporting links.The links can support high sounding repetition frequencies for Nyquist sampling while reducing report occupation on the sensing link.Source: A. Multi-Link Sensing · pp. 24–25
  • Multi-frequency sensing needs further study of non-coherent result fusion or coherent joint processing, alongside performance limits and algorithms under real Wi-Fi constraints.CSI feedback overhead grows with sensing bandwidth and may increase further across bands, motivating analysis of joint multi-band CSI compression.Source: A. Multi-Link Sensing · p. 25

B. Sensing with Artificial Intelligence (AI)

  • Potential standard-side AI applications include CSI feedback compression, adaptive sensing power control, and task-oriented feedback control that aims to minimize overhead while ensuring performance.The paper identifies GANs for compression and DRL, DNNs, and LSTMs as possible approaches to feedback control.Source: B. Sensing with Artificial… · pp. 25–26
  • AI-based sensing reports could support detection, recognition, estimation, radio-map reconstruction, environmental imaging, and applications combining WLAN sensing with communication.The paper describes these AI directions as being at an early stage and requiring further investigation.Source: B. Sensing with Artificial… · p. 26

C. Sensing at Intelligent Edge

  • Edge learning raises joint communication, computation, and sensing resource-management questions, including task-oriented design for heterogeneous devices and sensing goals.Federated edge learning is discussed as a way to address data-privacy concerns and support distributed learning at Wi-Fi sensing access points.Source: C. Sensing at Intelligent Edge · p. 26
  • Edge inference research needs to analyze performance, balance sensing-communication-computation trade-offs, adapt models to custom environments, and transmit task-oriented semantic information.Models that perform well on average across edge-device datasets may perform poorly in a specific environment; semantic transmission is proposed to ease split-inference communication bottlenecks.Source: C. Sensing at Intelligent Edge · pp. 26–27

D. Sensing with Intelligent Reconfigurable Surface (IRS)

The section outlines future directions spanning IRS-enabled sensing, compatibility with 6G ISAC, and integration of sensing with wireless power transfer. The paper concludes with an overview of IEEE 802.11bf standardization efforts, sensing procedures, candidate features, evaluation, and open challenges.

D. Sensing with Intelligent Reconfigurable Surface (IRS)

  • IRS-enabled WLAN sensing spans four architectures determined by whether reflecting elements are passive or active and whether the IRS has dedicated sensors.Active elements can amplify incident signals; dedicated sensors place target sensing at the IRS, while architectures without them place sensing at the AP.Source: D. Sensing with Intelligent… · p. 27
  • Transparent IRS sensing leaves the existing protocol unchanged, whereas non-transparent sensing treats the IRS as a cooperative node for joint signal and reflection design, with greater complexity.The cooperative design can achieve enhanced sensing performance, but requires more complex protocol design; challenges include inaccessible instantaneous AP–IRS CSI and joint processing of direct and reflected-link data.Source: D. Sensing with Intelligent… · p. 27

E. Compatible Technologies for ISAC and Wi-Fi Sensing

  • Shared techniques for 6G ISAC and WLAN sensing could reduce deployment and implementation costs and enable cooperative sensing between access points and small base stations.At 60 GHz and mmWave/THz, spherical-wave near-field propagation could support 3D localization without stringent synchronization requirements.Source: E. Compatible Technologies for ISAC… · p. 27

F. Wi-Fi Sensing and ISAC with Wireless Power Transfer (WPT)

  • Integrating sensing, communications, and wireless power transfer is presented as an emerging direction for next-generation Wi-Fi and future 6G.The paper describes sensing information as supporting more accurate beams toward IoT devices, alongside wireless powering and communication.Source: F. Wi-Fi Sensing and ISAC with… · p. 28

VIII. CONCLUSION

  • The paper surveys IEEE 802.11bf use cases and KPI requirements, sensing procedures, candidate features, evaluation methods, channel modeling, simulation, and future challenges.The authors hope the overview will provide insights for researchers and practitioners and promote technical advances and practical deployment.Source: §VIII · p. 28
Loading 2310.17661v1…