PaperlayerView exact revision on arXiv

Source-linked AI summary

Integrated Sensing and Communication in 3GPP: Standardization, Performance Evaluation, and Experimental Validation Toward 6G

Gosan Noh, Gebremeskel Seiley Abraha, Joshua Yabana, Nathaniel Foli, Michael Ankomah Asare, Evelyn Seyram Kakah, Mildred Nyantakyiwaa Okyere, De-Roy Ekow Sagoe, Shrayan Das, Pekka Kyösti, Junhyeong Kim, Kyoung-Jae Lee, Eunkyung Kim

arXiv:2610.22995v1eess.SP

TL;DR

Cellular ISAC’s standardization, performance evaluation, and target characterization are often treated separately. This article connects those views through a 3GPP progression review, simulation case studies, and UAV measurements; hybrid fusion improved detection, and practical predictive handover improved interruption and throughput outcomes.

  • Problem

    The paper addresses cellular ISAC across three views—3GPP standardization, reproducible performance evaluation, and measurement-based target characterization—that are often evaluated separately.

    Source: §VIII · p. 25

  • Method

    The article reviews 3GPP ISAC development, presents standards-aligned simulation case studies, and characterizes UAV radar cross sections through measurement.

    Source: §I · p. 2, §VIII · p. 26

  • Results

    Hybrid fusion improved detection with conditional estimation accuracy near hard-decision fusion, while practical predictive handover reduced interruption (95.3 to 58.0 ms) and increased throughput (14.50 to 15.79 Mbit/s).

    Source: §VIII · pp. 25–26

  • Takeaways & Limitations

    Measured UAV responses support modeling target visibility as a correlated process rather than as a fixed scattering coefficient.

    Source: §VIII · p. 26

    Sparse false-alarm events and primary handover failures do not support superiority claims for those outcomes.

    Source: §VIII · p. 26

Abstract

from arXiv · show

Integrated sensing and communication (ISAC) is evolving from a primarily physical-layer research into a standardized cellular-network capability spanning channel modeling, radio procedures, network architecture, and the delivery of sensing services. We review this progression within the Third Generation Partnership Project (3GPP), from Release 19 service requirements and target-aware channel modeling to the Release 20 New Radio (NR) sensing evaluation framework, radio access network procedures, and system architecture, and discuss future directions toward 6G. We then present two 3GPP-aligned simulation case studies to demonstrate the corresponding evaluation approaches. The first case study compares hard-decision, soft-decision, and hybrid fusion for unmanned aerial vehicle (UAV) sensing using multiple transmission and reception points (TRPs). Detection, false alarms, and conditional position and velocity accuracy are evaluated using receiver-derived angle estimates and separately calibrated method-specific thresholds. The second case study compares ideal and practical sensing-assisted predictive handover with conventional handover, while accounting for sensing error and information age. We further present a testbed-based measurement study that characterizes monostatic and bistatic UAV radar cross sections across frequencies, platforms, and observation angles and provides empirical support for correlated target-response modeling. Finally, we discuss open challenges and future research directions in multi-node fusion, resource and interference management, network evolution, security, privacy, and reproducible evaluation. Together, these components connect calibrated channel models, link- and system-level simulation, real-time prototypes, controlled measurements, and field trials through a staged evaluation roadmap.

I. INTRODUCTION

Cellular ISAC requires coordinated channel models, evaluation methods, radio procedures, network functions, and service mechanisms, not physical-layer waveform design alone. This review connects 3GPP’s evolving standards work with simulation, measurement, and a staged path toward deployment.

I. INTRODUCTION

  • ISAC reuses radio spectrum, waveforms, antenna arrays, and network infrastructure for sensing and data delivery across applications including UAV monitoring and transportation safety.Sensing can also support beam management, blockage prediction, and mobility, while connectivity coordinates distributed sensing.Source: §I · p. 1
  • Release 19 established service requirements and target-aware channel modeling, while Release 20 developed an NR sensing evaluation baseline, RAN sensing procedures, and service architecture.The review connects RAN1 evaluation, RAN3 realization, and SA2 service coordination.Source: §I · pp. 1–2
  • The article contributes a 3GPP cross-layer review, standards-aligned UAV-sensing and predictive-handover case studies, UAV target-response measurements, and a standards-to-deployment evaluation roadmap.The roadmap links standardized scenarios and requirements to calibrated channel and link models, system evaluation, prototypes, controlled measurements, and field trials.Source: §I · p. 2

A. Scope and Review Questions

  • The review asks how cellular sensing requirements can become implementable ISAC functions by jointly examining targets and channels, radio acquisition, network control and delivery, and evaluation.Its three questions cover 3GPP-specified functions and interfaces, simulation abstractions that separate algorithmic gains from operating-point changes, and measurements needed to validate simulation assumptions.Source: A. Scope and Review Questions · pp. 2–3

B. Literature Classification and Comparison Criteria

  • The review classifies studies by evaluation level, topology, scenario, channel basis, implementation maturity, and metrics across standards, simulations, prototypes or testbeds, and measurements.It compares numerical results only when assumptions and metric definitions are compatible; otherwise, comparisons remain qualitative.Source: B. Literature Classification and… · p. 3

C. Positioning Relative to Existing Surveys

This review connects existing ISAC survey perspectives to cellular standardization and evaluation, framing sensing as a network process that links target observations to decisions. It distinguishes sensing topologies and processing architectures because their geometry, synchronization, and information exchange shape evaluation.

C. Positioning Relative to Existing Surveys

  • This article connects Release 19 channel models and Release 20 evaluation assumptions with RAN3 procedures, SA2 architecture, simulation case studies, and UAV target-response measurements.Related surveys cover topics ranging from mobile-network sensing and fundamental limits to channel modeling, prototypes, and standards,,,,.Source: C. Positioning Relative to Existing… · p. 3

A. From Communication-Centric to Perceptive Cellular Networks

  • ISAC treats delay, Doppler, angle, amplitude, and their temporal evolution as observations of physical objects, linking channel assumptions to sensing outputs and network decisions.The target-induced path includes a UAV response that varies with frequency, geometry, polarization, and sensing direction.Source: §III · p. 3, Figure 2 · p. 4
  • Joint sensing and communication produces sensing and communication outputs, sensing-assisted communication uses environmental or motion estimates for network control, and communication-assisted sensing uses network coordination to improve sensing.A sensing estimate and a network outcome such as reduced handover interruption are distinct performance measures.Source: A. From Communication-Centric to… · p. 4
  • Stale or overconfident sensing information can degrade network operation, so evaluation should report both sensing quality and resulting communication performance.Radio-resource and beam decisions change the geometry of later observations, creating a feedback loop.Source: A. From Communication-Centric to… · p. 4, B. Sensing Topologies and Processing… · p. 5

B. Sensing Topologies and Processing Architectures

  • Monostatic sensing colocates transmitter and receiver but must account for weak echoes and leakage, whereas bistatic sensing separates them and adds two-leg delay, synchronization, and angle-dependent target response.Multistatic sensing can improve coverage and reduce geometric ambiguity when measurement association, error dependence, visibility, and common biases are handled appropriately.Source: B. Sensing Topologies and Processing… · p. 4
  • Network-centric processing transfers observations or intermediate results for infrastructure fusion, device-centric processing is local, and hybrid processing keeps time-critical work near the receiver.Higher-level reports reduce transport needs but embed more local processing decisions, including detection, association, filtering, and uncertainty estimation.Source: B. Sensing Topologies and Processing… · p. 5

C. Channel, Target, and Interference Abstractions

The section models sensing observations through target and background channels, geometry-dependent propagation, target scattering, and interference. These abstractions support receiver processing and network decisions that feed back into later radio choices.

C. Channel, Target, and Interference Abstractions

  • The sensing-receiver signal includes target and background propagation, residual self-interference, other-link interference, and receiver noise.The target and background MIMO channel responses preserve spatial and polarimetric information for array processing.Source: C. Channel, Target, and Interference… · p. 5
  • Bistatic delay follows the sum of transmitter-to-target and target-to-receiver distances, while Doppler reflects changes in total path length.For monostatic sensing, Doppler reduces to twice the radial-velocity projection divided by wavelength.Source: C. Channel, Target, and Interference… · p. 5
  • RCS scales target-path power and can vary with frequency, target class, aspect, bistatic angle, polarization, and time.UAV components scatter differently across aspect and frequency, so realistic responses may require target-dependent models rather than a fixed scalar.Source: C. Channel, Target, and Interference… · pp. 5–6
  • The cellular ISAC loop feeds link and object information into environment-aware control, coupling sensing quality to subsequent waveform, resource, and beam decisions.The communication and sensing branches share a radio front end.Source: Figure 3 · p. 6
  • Target visibility across TRPs may be correlated, and spatially or temporally correlated clutter can alter detector behavior compared with independent white noise.Shared blockage, target orientation, or large-scale environmental conditions can correlate links.Source: C. Channel, Target, and Interference… · p. 6
  • Interference includes residual self-interference, mutual sensing interference, and communication–sensing cross-interference, with effects modeled explicitly or by validated equivalent noise.External transmissions can also provide sensing illumination when waveform knowledge, synchronization, and calibration permit.Source: C. Channel, Target, and Interference… · p. 6

D. Evaluation Levels, Metrics, and Statistical Protocols

The section distinguishes complementary evaluation levels and specifies metrics and statistical practices for interpretable comparisons. It also connects communication–sensing tradeoffs with 3GPP’s staged standardization, from Release 19 service requirements to cellular sensing services.

D. Evaluation Levels, Metrics, and Statistical Protocols

  • Link-level, system-level, and end-to-end implementation evaluations capture different aspects, from waveform and detection detail to many-node behavior and implementation delay or interfaces.Results should be interpreted within the scope actually evaluated.Source: D. Evaluation Levels, Metrics, and… · p. 6
  • Detection and false-alarm probabilities require explicit thresholds, populations, association rules, and counting methods; conditional estimation errors should be reported alongside detection probability.Statistical comparisons should separate calibration from evaluation samples and use matched operating points and paired realizations where possible.Source: D. Evaluation Levels, Metrics, and… · pp. 6–7

E. Communication–Sensing Tradeoffs

  • Sensing quality competes with communication resources, latency, coverage, interference, and network load, so design seeks a feasible operating region rather than optimizing either function alone.More bandwidth, coherent observation time, aperture, or sensing-data exchange can improve sensing resolution or fusion while imposing communication or implementation costs.Source: E. Communication–Sensing Tradeoffs · p. 7
  • Network-level evaluation should represent the feedback between sensing observations and radio decisions, weighing communication loss avoided against sensing acquisition and use costs.Ideal sensing-assisted performance is an upper reference; practical methods must account for observation errors and information delay.Source: E. Communication–Sensing Tradeoffs · p. 7

IV. ISAC STANDARDIZATION IN 3GPP

  • 3GPP’s work spans Release 19 service requirements and target-aware channel modeling, followed by Release 20 radio evaluation, RAN procedures, and sensing-service architecture.Study reports assess feasibility, whereas normative specifications define interoperable behavior.Source: §IV · p. 7

A. Release 19 Foundations

  • Release 19 established service and channel-modeling foundations by identifying sensing use cases and requirements, deployment scenarios, sensing modes, and target-aware models.These activities established the basis for subsequent cellular ISAC work.Source: A. Release 19 Foundations · p. 7

1) Services and Use Cases:

  • Release 19 service requirements cover the process from authorized sensing requests and data collection to quality-qualified results delivered to applications under operator policy.Requirements also address privacy, consent, integrity, continuity, prioritization, and exposure; services may combine 3GPP observations with non-3GPP sources.Source: 1) Services and Use Cases · pp. 7–8

2) Deployment Scenarios and Sensing Modes:

Release 19 scenarios and sensing modes define the physical conditions that ISAC channel models must represent. The channel framework adds target-centered scattering and geometry while retaining background propagation and 3GPP scenario compatibility.

3) ISAC Channel Modeling:

  • Clause 7.9 separates target channels from background channels and represents targets through one or more scattering points with state and propagation attributes.The received response combines target-related paths with clutter and non-target reflections.Source: 3) ISAC Channel Modeling · p. 8

2) Deployment Scenarios and Sensing Modes:

  • Table IV groups deployment scenarios into UAV, human, automotive, industrial AGV, and hazardous-object sensing across varied targets, environments, and motion conditions.The scenarios include UMi, UMa, and RMa environments.Source: 2) Deployment Scenarios and Sensing… · p. 8, Table IV · p. 9
  • Six sensing modes vary whether TRPs or UEs transmit and receive and whether they are co-located; TRP–TRP bistatic sensing avoids receiver self-interference but requires inter-node synchronization.UE-involved modes add coverage and observation geometries while bringing capability, energy, authorization, mobility, and reporting considerations.Source: 2) Deployment Scenarios and Sensing… · p. 8, Figure 5 · p. 9

3) ISAC Channel Modeling:

  • Single-SPST models approximate compact targets, whereas multiple SPSTs represent extended targets whose RCS may depend on frequency, orientation, observation direction, and polarization.Spatial consistency lets delays, angles, powers, visibility, and RCS evolve continuously for tracking and sensing-assisted mobility evaluation.Source: 3) ISAC Channel Modeling · p. 8, Figure 6 · p. 9, Figure 7 · p. 10
  • Clause 7.9 defines channel generation, not an end-to-end sensing system; detector thresholds, multi-target association, multi-TRP fusion, and reporting-interface design remain outside its scope.Reproducible evaluations must additionally specify target, array, polarization, interference, receiver-processing, and statistical assumptions.Source: 3) ISAC Channel Modeling · pp. 9–10
  • The generation flow reuses the applicable TR 38.901 background procedure, builds geometry-consistent target paths, then superposes target and background coefficients at the receiver.This preserves compatibility with established propagation scenarios while adding object-centric sensing quantities.Source: 3) ISAC Channel Modeling · pp. 8–9, Figure 8 · p. 10

B. Release 20 NR Sensing Study

Release 20 extends Release 19’s target-aware channel foundation through a gNB-based UAV sensing evaluation framework, RAN procedures, and service architecture. The framework specifies evaluation assumptions and metrics, while the RAN interface supports sensing-task control and reporting at different abstraction levels.

B. Release 20 NR Sensing Study

  • Release 20 coordinates radio feasibility, RAN procedures, and core-network support, but its scope remains narrower than the Release 19 service vision.RAN1 established the gNB-based UAV sensing evaluation framework; RAN3 studied sensing control and information transfer, while SA2 studied service management and result delivery.Source: B. Release 20 NR Sensing Study · p. 10

1) Evaluation Scope and Assumptions:

  • The baseline workflow configures the scenario, generates target and background channels, processes receiver observations, optionally fuses results, and computes target-level metrics.The baseline uses UMa-AV, with 500 m inter-site distance, 4 or 4.9 GHz, and UAV heights of 25–300 m; configurations 1 and 2 differ mainly in maximum BS transmit power, 52 and 37 dBm.Source: 1) Evaluation Scope and Assumptions · p. 10, B. Release 20 NR Sensing Study · p. 10
  • Type-1 FAP measures target-absent drops with any detection, whereas Type-2 FAP averages unassociated detections in target-present drops.Their conditioning populations and denominators differ, so the metrics must be interpreted separately; the association rule also determines missed targets and estimates used for accuracy evaluation.Source: 1) Evaluation Scope and Assumptions · p. 10
  • The evaluation objectives are feasibility references, not service-level guarantees or evidence of coverage across other scenarios and impairments.End-to-end performance also depends on scheduling, reporting, processing and transport latency, fusion, tracking, and service exposure.Source: 1) Evaluation Scope and Assumptions · p. 10

2) Network Architecture and Interface:

  • The Ns interface connects the SenF and gNB for sensing control and information transfer, while the SenF can select gNBs and each gNB retains local radio-resource control.Fig. 9 combines the RAN and service-architecture views; Ns is the SenF–gNB-CU interface, SEN2 denotes the relationship in the SA2 reference-point model, and F1 connects the CU and DU.Source: 2) Network Architecture and Interface · p. 11, Figure 9 · p. 11
  • TR 38.765 recommends SCTP-based transport for sensing data in a potential normative phase and presents NsAP over SCTP/IP for sensing signaling.The study considered SCTP-, GTP-U/UDP-, and WebSocket-based alternatives for Ns exchanges.Source: 2) Network Architecture and Interface · p. 11
  • The Ns procedures cover interface setup and configuration, plus sensing initiation, modification, abort, and failure handling.Setup and configuration procedures establish or update the operational relationship; sensing procedures manage measurement requests and their lifecycle.Source: 2) Network Architecture and Interface · pp. 11–12
  • Levels A–D move from channel samples through profiles and detected parameters to object-level reports; Level C is a practical baseline for multi-gNB fusion.Level C avoids transporting raw samples or profiles, while Level D can reduce transport but embeds local processing decisions; interoperable fusion needs common semantics, coordinates, timing, confidence definitions, and quantization.Source: 2) Network Architecture and Interface · p. 12

3) Service Architecture and Procedures:

The Release 20 service architecture places the SenF between sensing-service consumers and gNB sensing entities, with reporting and authorization procedures for delivering results. The path toward normative work prioritizes interoperable gNB–SenF procedures and shared measurement semantics, while treating the current baseline as limited rather than comprehensive.

3) Service Architecture and Procedures:

  • The Release 20 Stage 2 baseline assigns a SenF to coordinate gNB-based aerial-object detection and tracking, with an untrusted AF accessing the service through the NEF.An operator-trusted AF can interact directly with the SenF through SEN3; an untrusted AF uses N33 to the NEF and SEN1 from the NEF to the SenF.Source: 3) Service Architecture and Procedures · p. 12
  • After authorization, the SenF selects sensing entities, obtains data or results, and delivers consolidated results according to one-time, periodic, or event-triggered reporting.Requests can specify a sensing time window, target area, reporting parameters, accuracy, missed-detection and false-alarm requirements, and service conditions.Source: 3) Service Architecture and Procedures · p. 12
  • The baseline is limited to gNB-based sensing in one PLMN, one SenF per request, and no dedicated sensing-data storage function.TS 23.137 is a draft Release 20 Stage 2 architecture specification, not a comprehensive architecture for all sensing modes and entity types.Source: 3) Service Architecture and Procedures · p. 12

C. From the Release 20 Baseline Toward 6G

1) Prospects for NR ISAC Normative Work:

  • Interoperable measurement reports require shared units, coordinates, dynamic ranges, timing, confidence, and target-association rules across gNBs and the SenF.The work item focuses on Levels C and D and quantization for the UAV use case.Source: 1) Prospects for NR ISAC Normative Work · p. 13
  • Authorization, privacy, integrity, confidentiality, auditability, and exposure policy must cover measurements and derived object results.Reporting control also needs freshness and confidence rules, conditions for event-triggering, and behavior when requested quality cannot be sustained.Source: 1) Prospects for NR ISAC Normative Work · p. 13
  • Release 20 study options are not guaranteed to become normative; the work item is a bridge toward interoperable implementation, not a complete ISAC specification.The normative phase selects a tractable subset of the alternatives evaluated in study work.Source: 1) Prospects for NR ISAC Normative Work · p. 13

2) Vision and Future Directions for 6G ISAC:

The section frames 6G ISAC as an expansion beyond the constrained NR baseline and reviews simulation abstractions for evaluating sensing and sensing-assisted communication. It emphasizes that evaluation choices must preserve the physical and network dynamics relevant to the question being studied.

2) Vision and Future Directions for 6G ISAC:

  • 6G study directions extend beyond gNB-monostatic UAV sensing to more target classes, sensing modes, waveforms, network roles, and resource-coordination needs.These are study directions and architectural questions, not committed normative features.Source: 2) Vision and Future Directions for 6G… · pp. 13–14

V. 3GPP-ALIGNED ISAC SIMULATION: METHODOLOGY, CASE STUDIES, AND RESULTS

  • The review presents two 3GPP-aligned case studies: multi-TRP UAV sensing at link level and sensing-assisted predictive handover at system level.Both use scenarios, channel models, antenna or link assumptions, and radio parameters drawn from relevant 3GPP documents.Source: §V · p. 14

1) Link-Level Sensing Evaluation:

  • Link-level sensing evaluation preserves delays, angles, Doppler, phases, and time evolution needed for target detection and parameter estimation.Bistatic and multistatic evaluations also need two-leg propagation, node geometry, target scattering, and common time and frequency references.Source: 1) Link-Level Sensing Evaluation · p. 14

2) System-Level ISAC Evaluation:

  • System-level simulation represents networks and mobility over extended time using simplified waveform and receiver models calibrated to physical-layer behavior.For mobility evaluation, sensing information’s update period, age, uncertainty, missed detections, and possible correlated biases matter alongside protocol states.Source: 2) System-Level ISAC Evaluation · pp. 14–15
  • Simulation abstraction should follow the evaluation objective, distinguishing standards-derived parameters, link-level calibration, and study-specific policy assumptions.Resource allocation may use analytical sensing-SINR constraints, whereas mobility evaluation needs time-evolving measurements and a protocol state machine.Source: 2) System-Level ISAC Evaluation · p. 15

3) Networked and Multi-TRP Sensing:

The multi-TRP case study compares calibrated hard-, soft-, and hybrid-fusion methods using receiver-derived angles and target-directed observations. Hybrid fusion detects the most targets while retaining conditional estimation accuracy close to hard-decision fusion, but false-alarm differences are inconclusive and the evaluation is limited in scope.

1) Scenario, Channel, and Compared Fusion Methods:

  • The study uses a 3GPP-aligned UAV setup with three TRPs in a study-specific triangular geometry, target-directed observations, and receiver-derived angle estimates.It includes residual self-interference but omits inter-TRP interference; the results are not from blind spatial search or continuous tracking.Source: 1) Scenario, Channel, and Compared… · p. 15
  • The three methods use one-to-one association, with method-specific thresholds calibrated to a common 1% Type-1 FAP budget and frozen before evaluation.The evaluation combines 100 target-present scenes, 500 target opportunities, and 784 target-absent scenes; position and velocity CDFs include only correctly detected and associated targets.Source: 1) Scenario, Channel, and Compared… · p. 15

2) Multi-TRP Detection and Estimation Results:

  • Hybrid fusion detected 418/500 targets (0.836), versus 398/500 (0.796) for hard-decision and 359/500 (0.718) for soft-decision fusion.Hybrid and soft-decision Type-1 FAP were both 10/784 (1.276%), versus 7/784 (0.893%) for hard-decision; Type-2 FAPs were 0.367%, 1.167%, and 0.867%, respectively.Source: 2) Multi-TRP Detection and Estimation… · p. 15
  • The 90th-percentile 3D position errors were 2.897, 3.260, and 2.986 m for hard-, soft-, and hybrid-decision fusion, respectively.Corresponding horizontal-velocity errors were 1.302, 3.756, and 1.513 m/s; hybrid retains accuracy close to hard-decision fusion, but missed-detection probabilities remain above the 5% reference objective.Source: 2) Multi-TRP Detection and Estimation… · pp. 15–16, 3) Key Findings and Evaluation Scope · p. 16

3) Key Findings and Evaluation Scope:

  • The findings are limited to controlled multi-link link-level evaluation; continuous tracking, blind spatial search, and communication–sensing coexistence are outside scope.Conditional estimation CDFs should be interpreted together with detection and false-alarm probabilities.Source: 3) Key Findings and Evaluation Scope · p. 16

D. Case Study B: Sensing-Assisted Predictive Handover

The case study evaluates predictive handover under shared execution conditions, comparing conventional, ideal, and practical sensing-assisted methods while accounting for sensing uncertainty and information age. Practical prediction reduces interruption and increases throughput relative to conventional handover, but the small failure count limits reliability conclusions.

1) Scenario, Mobility, and Compared Handover Methods:

  • The study models a UE at 60 km/h crossing a three-TRP outdoor UMa deployment; Practical prediction uses noisy position and velocity estimates and information aged 0.1–0.5 s.All methods share a 3 dB RSRP advantage and 3 s persistence requirement; predictive preparation uses a 2 s look-ahead and assumes successful preparation reduces modeled access time from 60 to 30 ms.Source: 1) Scenario, Mobility, and Compared… · p. 16

2) Handover and User-Throughput Results:

3) Key Findings and Evaluation Scope:

VI. ISAC PROTOTYPES, TESTBEDS, MEASUREMENTS, AND EXPERIMENTAL CASE STUDY

  • The study-specific scalar-link simulation cannot establish implementation-chain or physical target-response realizability, motivating complementary prototypes, testbeds, and measurements.The measurements connect target visibility to the TR 38.901 Clause 7.9 channel abstraction.Source: §VI · p. 18

A. Prototype and Testbed Literature

The section distinguishes prototype and networked-system validation from channel and target-response measurement, then describes a UAV RCS study designed to supply measured target-response inputs for ISAC modeling.

1) Waveform and Signal-Processing Prototypes:

  • ISAC prototypes commonly share a radio front end and multicarrier waveform, with communication and sensing processing branches; SDRs, FPGAs, GPUs, or accelerators support implementation and real-time processing.Validation should report communication and sensing metrics under the same operating configuration, rather than select best results from different resource allocations.Source: 1) Waveform and Signal-Processing… · p. 18

2) Networked ISAC Demonstrations:

  • Networked demonstrations test whether sensing information survives practical synchronization, calibration, transport, latency, and delivery constraints, but multi-node radio operation alone does not validate an end-to-end service.Service validation also covers task authorization and management, information transfer and delivery, latency, and failure handling.Source: 2) Networked ISAC Demonstrations · p. 18

B. Channel and Target Measurement Literature

  • Channel campaigns characterize propagation, while target-response campaigns measure object scattering across frequency, incidence, observation direction, polarization, and target state; combining them without distinction limits transferability.UAV response varies with platform, aspect, and observation angle, and static airframe measurements do not capture rotor micro-Doppler or time variation.Source: B. Channel and Target Measurement… · pp. 18–19

C. Experimental Case Study: UAV RCS Characterization

  • The case study measures UAV target scattering rather than end-to-end sensing performance, providing empirical inputs to parameterize and validate RCS models in TR 38.901 and the TR 38.765 UAV scenario.The measurements target dependence on frequency, observation aspect, platform geometry, and bistatic angle.Source: C. Experimental Case Study: UAV RCS… · p. 19

1) Measurement Scenario:

  • The principal Phantom 2 campaign scans azimuth in 1° increments and measures 2001 frequency points from 11 to 26 GHz; additional platform datasets support descriptive, not controlled, cross-platform comparison.The Phantom 2 bistatic configuration varies the bistatic angle from 0° to 90° in 15° increments; platforms, frequency ranges, and facilities differ.Source: 1) Measurement Scenario · p. 19

2) Testbed Implementation and Measurement Processing:

The measurement pipeline extracts calibrated UAV RCS samples, revealing frequency-, aspect-, platform-, and geometry-dependent responses. These findings motivate correlated target-response models and define the limits of the campaign’s coverage.

2) Testbed Implementation and Measurement Processing:

  • The processing chain coherently subtracts background, gates the target return in delay, processes frequency subbands, and normalizes against a reference cylinder to estimate absolute RCS.The reference ratio removes common gain and propagation terms to first order, but alignment, gate selection, near-field behavior, and residual background remain sources of uncertainty.Source: 2) Testbed Implementation and… · p. 19

3) Measurement Results:

  • Measured UAV RCS varies jointly with frequency, aspect, platform, and observation geometry; monostatic values are generally higher in the sampled geometries, and angular changes can sharply weaken visibility.The monostatic–bistatic ordering is not universal, and frequency responses include local peaks and notches; the frequency-averaged azimuthal RCS can be modeled log-normally with platform- and band-dependent parameters.Source: 3) Measurement Results · p. 19

4) Implications for 3GPP-Oriented ISAC Evaluation:

  • Target class and orientation should shape RCS, while angular visibility should be modeled as correlated; coverage-level evaluation may use averages, but detection, tracking, and multi-node evaluation should retain angular fluctuations.The measurements also support target-induced path-loss models in which class and frequency affect average loss while bistatic visibility remains angle-dependent.Source: 4) Implications for 3GPP-Oriented ISAC… · p. 22
  • Monostatic-to-bistatic equivalence is safest in restricted geometries; direct bistatic measurements or calibrated stochastic models are preferable when target complexity violates model assumptions.The figure compares equivalence-model predictions with measured bistatic RCS at 11 and 26 GHz and notes increased mismatch for some wideband, large-angle conditions.Source: 4) Implications for 3GPP-Oriented ISAC… · p. 22
  • The campaign omits full polarimetry, rotor micro-Doppler, translational motion, attitude dynamics, and systematic elevation dependence; Phantom 2 measurements use a controlled non-anechoic near-field setup.These boundaries constrain interpretation of the extracted RCS statistics.Source: 4) Implications for 3GPP-Oriented ISAC… · p. 22

VII. OPEN CHALLENGES AND PATH TOWARD DEPLOYMENT

  • The measurements provide target-response inputs to the broader validation roadmap, while the open-challenges section identifies five coupled technical challenges for deployment.The roadmap context is that measured inputs complement, rather than replace, other evaluation stages.Source: 4) Implications for 3GPP-Oriented ISAC… · p. 22, §VII · p. 22

A. Channel and Target-Response Generalization

The section identifies challenges in generalizing target-response models, coordinating multi-node sensing, balancing radio resources, and evolving RAN and service architectures toward scalable deployment.

A. Channel and Target-Response Generalization

  • Target-response models calibrated on limited platforms and conditions must generalize across target classes, structural variants, motion, sensing modes, and environments; TR 38.901 provides structure but not universal parameters.A hierarchical model combines class distributions, persistent platform properties, correlated orientation- and motion-dependent processes, deterministic object responses, and stochastic residuals.Source: A. Channel and Target-Response… · p. 22, Figure 26 · p. 23

B. Multi-Node Coordination, Fusion, and Tracking

  • Multi-TRP fusion can underperform a reliable single link when reports are misassociated or correlated errors are treated as independent; interoperable operation needs shared references, identifiers, lifecycle rules, and failure handling.The evaluation agenda includes correlated visibility, clutter, report loss, and delay, followed by staged validation from common-clock setups to independently operated sites.Source: B. Multi-Node Coordination, Fusion,… · pp. 22–24

C. Resource Allocation, Waveforms, and Interference Management

  • Sensing and communication compete for finite radio resources, so allocation must balance sensing performance against communication quality of service and implementation cost; waveform, reference-signal, and resource design are coupled.The section also calls for managing self-interference, mutual sensing interference, and sensing–communication cross-interference, with target-aware beam selection distinct from communication beam selection.Source: C. Resource Allocation, Waveforms, and… · p. 24

D. RAN and Core-Network Evolution

  • Scaling Release 20’s initial architecture requires clear RAN and SenF ownership of configuration, processing state, report identity, and recovery, plus service continuity across entities and network functions.The proposed validation uses interface-in-the-loop tests with varied transport, processing, and failure conditions; a hierarchical SenF split is organized around latency, transport, privacy, scalability, and failure containment.Source: D. RAN and Core-Network Evolution · p. 24

E. Security, Privacy, and Trustworthy Sensing

The section highlights security and privacy risks in cellular sensing, outlines a flexible reproducibility roadmap, and summarizes simulation and measurement findings that bound how cellular ISAC performance should be interpreted.

E. Security, Privacy, and Trustworthy Sensing

  • Cellular sensing can expose sensitive information about non-subscribers, while spoofing, replay, reference manipulation, corrupted tracks, and sensing-result inference are identified threats.A high-confidence report is not itself a security guarantee, and object-level reporting alone does not ensure privacy because trajectories can reveal identity or behavior.Source: E. Security, Privacy, and Trustworthy… · pp. 24–25

F. Reproducible Evaluation Roadmap

  • The six-stage evaluation roadmap increases model scope, implementation detail, and operational realism without prescribing a fixed sequence.It distinguishes research validation from conformance: conformance requires a normative test procedure with specified configurations, measurements, and acceptance criteria.Source: F. Reproducible Evaluation Roadmap · p. 25

VIII. CONCLUSION

  • Practical sensing assistance reduced mean handover-attributable interruption from 95.3 to 58.0 ms and increased mean user throughput from 14.50 to 15.79 Mbit/s versus conventional handover; sparse failures do not support a general reliability-superiority claim.Hybrid fusion improved detection while retaining conditional estimation accuracy close to hard-decision fusion, but sparse false alarms limited conclusions about false-alarm superiority.Source: §VIII · pp. 25–26
  • Monostatic and bistatic UAV measurements show RCS varies with frequency, platform, and observation aspect, with correlation over only a limited angular interval.The campaign used fixed polarization and controlled turntable rotation; extensions across polarization, elevation, bistatic angle, and target motion could support a dynamic polarimetric target response.Source: §VIII · p. 26
  • The paper frames cellular ISAC performance through the adopted model resolution, operating conditions, and validation level, while broader targets and network-wide sensing remain directions toward 6G.Its current baseline is gNB-based UAV sensing; broader target classes, UE participation, and multistatic cooperation remain study and evolution directions.Source: §VIII · pp. 25–26

Related papers

Loading 2610.22995v1…