Joint Communication and Sensing: 6G Integration

Updated 15 November 2025

Joint Communication and Sensing is a paradigm that integrates data transmission and radar-like sensing using shared waveforms and hardware for efficient 6G networks.
It leverages unified signal models and multi-domain waveform designs to optimize both spectral efficiency and precise environment mapping through resource sharing.
Practical architectures deploy massive MIMO, cloud-RAN, and AI-driven scheduling to facilitate autonomous vehicles, digital twins, and real-time robotics with high precision.

Joint Communication and Sensing (JCAS) refers to the integration of wireless communication and sensing functionalities—typically radar-like environment monitoring—onto a shared hardware and spectral resource pool. In JCAS, the same waveforms, antennas, time-frequency resources, and often the same baseband/RF chains, simultaneously serve high-throughput data transmission and radio-based sensing tasks such as localization, imaging, or environment mapping. The paradigm is a cornerstone for 6G networks, enabling advanced use cases like autonomous vehicles, real-time robotics, digital twins, and situational awareness, while achieving high spectral- and hardware-efficiency by design (Wymeersch et al., 14 Feb 2024).

1. Theoretical Foundations and System Models

JCAS enables spectrum, hardware, and energy efficiency by co-designing the signal models and processing pipelines of communications and radar-style sensing. The principle is that both communication and radar are fundamentally parameter estimation problems performed over wireless channels, typically under linear time-varying multi-path models.

Unified Signal Models

Transmit waveform: Generic form

$x(t) = \sum_{m, n} s_{m,n} e^{j2\pi n\Delta f(t-mT)} \mathrm{rect}\left(\frac{t - mT}{T}\right)$

where $s_{m,n}$ encodes data and possibly dedicated sensing pilots (Wymeersch et al., 14 Feb 2024).

Sensing/echo model:

$y_{\text{sense}}(t) = \sum_{\ell=0}^{L-1}\alpha_\ell x(t-\tau_\ell)e^{j2\pi \nu_\ell t} + w(t)$

with delay $\tau_\ell$ , Doppler $\nu_\ell$ , and complex gain $\alpha_\ell$ .

Physical Layer Integration:
- Shared MIMO/OFDM/OTFS or SC-IFDM waveforms.
- Dual application of the same physical-layer resource for data and parametric environment extraction.
Key trade-off: The spectral footprint of communications and radar are nearly additive, forming an approximately zero-sum game for bandwidth, but not for power (Li, 2020).

2. JCAS Waveform and Architecture Design

The waveform and resource allocation design space is highly active, with the primary axis being communications-centric, sensing-centric, or truly joint/synergetic designs (Zhang et al., 2021, Wymeersch et al., 14 Feb 2024).

Exemplary Joint Waveform Designs

A. SC-IFDM–FMCW Orthogonal Waveform (Boudjelal et al., 16 Mar 2025)

Combines single-carrier interleaved frequency division multiplexing (SC-IFDM, a 5G candidate) and frequency-modulated continuous wave (FMCW) radar chirp in the DFT domain.
Data and chirp samples are mapped into orthogonal "slots" in a 2D DFT grid:

$X^{\mathrm{comb}}(k,l) = \begin{cases} \sqrt\psi\, s^{\mathrm{FMCW}}(l)\,e^{j2\pi(-kl/(MN))}, & [\tfrac M2 +l-k]_N=0\ X^{\mathrm{SC-IFDM}}(k,l), & \text{otherwise} \end{cases}$

Enables perfect DFT-domain orthogonality and minimal data-sensing interference. The composite waveform is created via $MN$ -IDFT and cyclic prefix, with phase-shifts to ensure chirp continuity between frames.

B. Multi-Carrier MIMO JCAS with Subcarrier/DoF Partition (Nguyen et al., 2023)

Assigns only a subset of subcarriers to joint sensing-and-communication, optimizing beamformers via Riemannian manifold optimization under beampattern and throughput constraints.
Proposed approach yields 60% communication rate gains (at 10 dB SNR, $N_{\mathrm{JCAS}}=16$ out of 64 subcarriers) at no loss in sensing performance.

C. Dual-domain (FT-DD) and Code-Division Approaches (Rinaldi et al., 2021, Chen et al., 2023)

Superposition of an OFDM grid (communications in frequency–time) with a sparse delay–Doppler domain sensing signal.
Orthogonal code-division multiplexing across subcarriers enables post-processing gain for both data detection and radar estimation, offering up to $\sim$ 30 dB BER-equivalent improvement at low SINR.

Architectural Enablers

Massive/distributed MIMO arrays for angular/range–Doppler parameter resolution (Fang et al., 2022).
Cloud-RAN topologies for centralized fusion of raw I/Q or feature-compressed data (Zhang et al., 2017, Wymeersch et al., 14 Feb 2024).
Reconfigurable Intelligent Surfaces (RIS) to simultaneously enhance both comm and radar channels via spatial reconfiguration.
Unified protocol stacks and logical planes: The Sensing Management/Processing Functions (SeMF/SPF) orchestrate physical, networking, and data plane resource allocation, exposing environment data to higher-layer applications (Wymeersch et al., 14 Feb 2024).

3. Joint Resource Allocation, Performance Metrics, and Trade-Offs

Resource sharing between communication and sensing introduces multi-objective optimization.

Bandwidth and Power Budgeting

Bandwidth: JCAS is fundamentally bandwidth-limited; $B_c + B_s \approx B_{\mathrm{total}}$ . Rate and sensing accuracy scale according to

$R_c = B_c\log_2\left(1 + \frac{P_t G}{N_0 B_c}\right),\quad R_s \propto P_t B_s^3,\quad \text{(sensing: CRLB $\propto 1/(P_t B_s^3)$)}$

(Li, 2020).

Power: Power allocation is marginally conflictual, as spectrum partition primarily determines achievable operating points.
Optimization Problem (sample form (Wymeersch et al., 14 Feb 2024)):

$\min_{\{p_{c},p_{s},b_{c},b_{s}\}} -R(p_{c},b_{c}) + \lambda\,\mathrm{MSE}_\tau(p_{s},b_{s})$

subject to total power and bandwidth bounds.

Metric Table

Metric	Communication Aspect	Sensing Aspect
Data rate	Spectral efficiency (b/s/Hz)	Range/velocity estimation RMSE
SINR	Post-equalization, per carrier	Beat SNR, detection probability
Resolution	Bandwidth per stream	$c/2B$ (range), $\lambda/2T$ (velocity)
Pilot power/placement	Channel estimation, pilot design	Radar pilot (or combined) locations
Delay	Transmission/retransmission	Beam-training, beat-processing latency

Performance trade-offs are design-dependent:

More bandwidth or time for sensing increases physical–environment resolution but reduces data throughput.
Orthogonal waveform designs and shared pilot approaches mitigate overhead, enabling operation near the single-modality Pareto frontiers in both domains (Boudjelal et al., 16 Mar 2025).

4. Signal Processing and Inference Algorithms

Signal processing is grounded in multi-dimensional harmonic retrieval (e.g., delay, Doppler, angle), compressed sensing, and message-passing inference.

FFT-based and MUSIC-based estimation: FFT methods yield range/velocity granularity of $1/B$ and $1/T$, whereas subspace approaches (e.g., 2D MUSIC) afford super-resolution, at the cost of cubic computational complexity (Chen et al., 2022).
Compressed Sensing (CS): Utilized in scenarios exploiting the sparsity of environment or code domain; e.g., GAMP-based inference for scene reconstruction with sparse codebooks (Tong et al., 2021, Zhang et al., 2017).
Iterative Joint Detection: Alternating or sliding-window schemes transfer information between iterative communication decoding and environmental estimation, stabilizing performance in the presence of unknowns and channel non-stationarity.

5. JCAS in Practice: Architectures, Applications, and Implementation

Real-world Integration Scenarios

Vehicular autonomy (V2X): JCAS enables tape-measure positioning and velocity estimation while maintaining URLLC-grade connectivity (Wymeersch et al., 14 Feb 2024).
Smart manufacturing / digital twins: High refresh-rate environment mapping, with JCAS yielding $\leq$ cm-level accuracy at sub-ms update intervals (Feng et al., 2023).
XR/THz communications: Multi-GHz bandwidths allow joint sub-centimeter ranging and Tbps communication; reliable beam-tracking is achieved by feeding environment estimates into comm protocols (Chaccour et al., 2021).

Deployment and Compute Considerations

Edge/cloud fusion: Raw I/Q rates are often in the multi-Gb/s range, requiring distributed or hierarchical processing to meet latency and data regulations (Wymeersch et al., 14 Feb 2024).
AI/ML for scheduling: Adaptive resource allocation leverages ML or DRL-based controllers, optimizing for non-stationary user/sensing traffic (Feng et al., 2023).

Implementation Details

Waveform parameterization: E.g., for SC-IFDM–FMCW, typical blocks: $M=216$ , $N=16$ , $L_{cp}=16$ , $B_c=200\,$ MHz, $T_c=8.64\,\mu$ s.
Processing chain:
- At the receiver, radar: mix echo with chirp, 2D FFT for range–Doppler.
- At the receiver, comm: extract data subcarriers, use chirp slots as pilots for channel estimation.
Practical performance: For SC-IFDM–FMCW, range–Doppler mapping is unambiguous up to 80 m and $\pm70\,$ m/s; BER loss at data pilot ratios $\psi/\sigma_d^2=10…20$ dB is negligible compared to pure SC-IFDM (Boudjelal et al., 16 Mar 2025).

6. Future Directions and Open Problems

Distributed/large-scale MIMO and RIS: Realizing real-time, multi-modal environment mapping and spatial coverage (Wymeersch et al., 14 Feb 2024, Fang et al., 2022).
AI-native JCAS stacks: Deep learning models for joint waveform adaptation, resource allocation, and closed-loop environment-aware protocol stacks.
Synchronization and calibration: Sub-ns ranging demands tight time/frequency control; practical JCAS solutions must handle real-world clock and hardware impairments (Pegoraro et al., 2023).
Security and privacy: JCAS leaks contextual environment information; secure waveform/beam design and network-side control are required (Günlü et al., 2022).
Integration with other modalities: Cooperative fusion with LIDAR, camera, and acoustic sensors for robust SLAM and digital-twin applications (Wymeersch et al., 14 Feb 2024).
Energy/sustainability: Joint metric frameworks such as energy efficiency ratio (EER) for dual-function stacks (Wymeersch et al., 14 Feb 2024).

7. Summary Table of Joint Communication and Sensing Paradigms

Paradigm	Resource Sharing	Signal Model	Key Technical Challenge	Numerical Achievement
Orthogonal slot/waveform	Time/freq/pilot	OFDM/OTFS/SC-IFDM	Overhead, non-orthogonality	BER/RMSE within 0.5 dB of standalone
Full overlap	Code/pilot	CD-OFDM, dual-domain	Interference, SIC, power allocation	$30$\,dB BER/RMSE gain at low SINR(Chen et al., 2023)
Subcarrier partitioning	Frequency	Multi-carrier MIMO	Integer/nonconvex optimization	60% throughput gain at same RMSE(Nguyen et al., 2023)
Edge AI scheduling	Dynamic	Mixed	Model/observation uncertainty	30-50% latency reduction(Feng et al., 2023)

JCAS now underpins the architectural vision for 6G, targeting spectrum-coefficient, AI-driven, and environment-aware wireless connectivity, moving beyond decoupled, application-specific design toward network-native multimodal perception and interaction (Wymeersch et al., 14 Feb 2024, Boudjelal et al., 16 Mar 2025).