Integrated sensing and communication (ISAC) is a crucial technology in the future sixth-generation (6G) mobile communication network. Radio frequency (RF)-ISAC has been a major focus in recent years, but optical wireless (OW)-ISAC has also attracted significant attention due to its potential benefits, which include the enhancement of communication rates, the improvement of sensing precision, and the reduction of interference. In this article, we propose an OW-ISAC framework combining arbitrary waveform frequency modulated continuous wave (AW-FMCW) sensing and communication technology and multi-input multi-output (MIMO) technology. Leveraging the arbitrary waveform capability of AW-FMCW technology, joint sensing and communication design are achieved through frequency-modulated waveform encoding. MIMO technology, widely adopted by RF-ISAC, is being applied for the first time in OW-ISAC to achieve high-precision estimation of direction of arrival (DOA). At the transmitter end, the multi-beam property of optical phase array (OPA) is utilized for multi-target communication and sensing tasks. At the receiver end, a well-designed uniform linear photodiode (PD) array is employed to extract echo signals from a specific direction while eliminating interference from other signals. Furthermore, the decoding matrices and PD orientations are optimized to achieve adaptive directional sensing. Finally, extension numerical simulations are conducted to validate the performance of multi-channel integrated sensing and communication capabilities.
In this paper, we propose a joint multi-dimensional index modulation (JMD-IM) framework for orthogonal time frequency space (OTFS) to enhance the spectral efficiency (SE) and reliability in high-mobility channels. Unlike conventional multi-dimensional index modulation (MD-IM) schemes that typically perform IM independently across different domains, which may leave available indexing patterns unused or impose constraints on the number of indexing entities (e.g., requiring integer powers of 2), the proposed framework jointly utilizes indexing opportunities across delay-Doppler and space domains. This unified approach overcomes the limitations of existing designs and can be regarded as a generalization of conventional MD-IM techniques. Specifically, three schemes are introduced according to different spatial indexing granularities: block-level JMD-IM (B-JMD-IM), where each OTFS subblock employs one spatial index; symbol-level JMD-IM (S-JMD-IM), where each symbol within the subblock is spatially indexed; and in-phase and quadrature-level JMD-IM (IQ-JMD-IM), where the I/Q components of each symbol are independently indexed across antennas. A unified mathematical model is developed in the space-delay-Doppler domain, together with corresponding subblock-wise maximum-likelihood (ML) detector. Furthermore, we propose two algorithms to enhance system performance: 1) an index pattern optimization method, which improves reliability by enlarging the average Euclidean distance; 2) an adaptive low-complexity detector, which achieves promising performance with reduced computational cost. The theoretical expressions for SE and bit error probability (BEP) are derived. Extensive computer simulations confirm that the proposed JMD-IM schemes outperform existing MD-IM variants and conventional MIMO-OTFS/OFDM systems.
Adapting deep-learning receivers after deployment is challenging because ground-truth channel state and transmitted payload symbols are unavailable, while sparse pilots alone provide insufficient supervision. The orthogonal frequency-division multiplexing waveform, however, embeds physical structure that can provide label-free supervision through physics-informed constraints. Here, we advance Physics-Embedded Inverse Learning (PEIL) toward deployment in two phases. Phase I develops Improved PEIL, a compact receiver with 0.093M trainable parameters that coordinates carrier frequency offset estimation and channel state information refinement through symbol-recovery supervision. In Phase II, Physics-Embedded Label-free Adaptation (PELA) adapts the same receiver using pilot reconstruction together with cyclic-prefix phase consistency and delay–symbol smoothness of the channel impulse response tail. Trained through 16-QAM symbol recovery, Improved PEIL generalizes without retraining to unseen QPSK and 64-QAM payloads. Under the EVA-to-ETU transition at 35 dB, PELA reduces symbol error rate from 0.1501 to 0.0258 and block error rate from 0.7106 to 0.0591 relative to static Improved PEIL. Across the evaluated mobility shifts, PELA becomes beneficial when deployment conditions differ substantially from the training environment. These results show that protocol-native pilots and waveform physics can supply the missing supervision for selective post-deployment adaptation.
Radio-based simultaneous localization and mapping (radio-SLAM) offers a promising solution for user equipment (UE) localization and radio feature map construction without external positioning infrastructure, with its feasibility already demonstrated through integrated sensing and communication (ISAC) prototypes. However, the performance of radio-SLAM is constrained by the sensing capability of wireless devices and irregular UE motion, necessitating the fusion of multi-modal sensors to harness complementary advantages in multi-modal SLAM. Critically, existing multi-modal SLAM methods are underdeveloped and do not leverage SLAM outputs for communication enhancement, nor have they been demonstrated in prototypes. To address these challenges, we propose a multi-modal SLAM framework integrated with SLAM-aided beam management, fusing radio measurements, inertial measurement unit (IMU) data, and stereo camera images to enhance localization robustness and mapping accuracy in both line-of-sight (LoS) and non-line-of-sight (NLoS) scenarios under varying lighting conditions. The resulting localization and mapping information facilitates beam tracking and blockage prediction. We further develop an ISAC prototype system capable of collecting synchronized multi-modal data in dynamic scenarios. Experiments demonstrate that the proposed SLAM achieves decimeter-level UE localization and radio map construction under both bright and low-light conditions. The beam management module enables precise beam alignment with an average beam direction angular error of 0.0403 rad and blockage prediction with an average detection rate of 84.33% across LoS and NLoS paths. The collected dataset and implementation are publicly released to support reproducible research in multi-modal ISAC.
The integrated geostationary Earth orbit (GEO)-multibeam and low Earth orbit (LEO)-uncrewed aerial vehicle (UAV) network has emerged as a promising paradigm to enhance the coverage and capacity of terrestrial networks. However, due to the high dynamics of the network and the heterogeneous service demands of user equipments (UEs), joint handoff control and resource block (RB) allocation becomes a critical yet challenging problem. To address this issue, we propose a two-stage dynamic optimization framework, aiming to maximize throughput while minimizing energy consumption and handoff cost. Specifically, the original problem is decomposed into two subproblems: handoff control and RB allocation with fixed handoff decisions. Multi-agent dueling double deep Q-network (MAD3QN) is designed for handoff control, where the centralized training with decentralized execution (CTDE) method is utilized to share intelligent information, and an offline training approach is adopted to support practical deployment. Matching theory (MT) is then applied to determine RB allocation by finding a stable matching between RBs and UEs. Simulation experiments with real-world satellite deployments demonstrate that the proposed algorithm combining MAD3QN and MT effectively converges and significantly outperforms the existing baselines.
Orthogonal time frequency space (OTFS) modulation exploits the delay-Doppler (DD) domain to convert time-varying channels into quasi-static representations. Rectangul-arwaveform-based OTFS (RW-OTFS), as a practical implementation, is widely believed to inherit this advantage. However, in this paper, we demonstrate that in RW-OTFS, fractional Doppler shifts disrupt this time-invariance, leading to time-varying channel response distortions that challenge conventional channel estimation and detection methods. To enable reliable detection under time-variant DD domain channels, this work introduces a decision feedback-aided dynamic phase compensation (DFDPC) scheme for RW-OTFS. The proposed DFDPC scheme achieves high-precision estimation of fractional Doppler shifts for the time-varying DD domain channel and eliminates the cumulative effect of residual estimation errors through decision feedback from detected data symbols. Furthermore, the dynamic phase compensation in DFDPC also addresses estimation errors induced by fractional delays. Simulation results verify that DFDPC maintains superior performance under time-variant DD domain channels, while enabling reliable transmission with extended frame lengths.
Random multiplexing (RM) with approximate message passing (AMP)-type detectors effectively address doubly-selective fading in high-mobility scenarios while achieving maximum a posteriori (MAP) bit-error rate performance. To further improve RM performance, this paper considers oversampling at an integer multiple of the symbol rate. Traditional oversampling detectors are challenged by the high complexity of processing large-dimension stacked sampled data and substantial overhead from correlated noise whitening. To overcome these issues, we present the oversampled RM system model, treating each oversampling instance as a distributed node and modeling the system as a distributed detection problem. In this context, we propose a cooperative distributed AMP-type detection framework with inner-iterative local estimation and outer-iterative global fusion. Then, we analyze the correlation among the local messages caused by the shared RM matrix and the oversampled noise. The orthogonality of the estimation errors ensures that the matrix formed by local information from distributed nodes has columns that are independent and identically distributed, while each row follows a joint Gaussian distribution. By leveraging this correlation, we derive an optimal fusion strategy to design a low-complexity cooperative distributed memory AMP detector and a high-stability cooperative distributed orthogonal AMP detector. Our simulation results show that the proposed detectors outperform both the AMP-type detectors with Nyquist sampling and the oversampled centralized detectors without noise whitening. The performance of RM is superior to that of other multiplexing schemes when the same detector is employed. In the outer iteration, our optimal fusion outperforms the Gaussian fusion that does not consider correlation. Moreover, with sufficient inner iterations, a single outer iteration matches the performance of full outer iterations, minimizing communication overhead.
The advent of massive antenna arrays and intelligent surfaces in sixth-generation (6G) wireless networks opens an exciting frontier, where communication technologies in the near-field (NF) region will play a pivotal role. Given the intrinsic spatial correlation of NF channels with device and object locations, integrated sensing and communication (ISAC) emerges as a key enabler for realizing more context-aware and efficient NF communication systems. However, the evolution of ISAC also introduces a novel security challenge, which is the exposure of transmitter location information through adversarial sensing. Against this backdrop, this paper develops a twofold NF secure transmission framework to jointly protect communication confidentiality and transmitter-location privacy. Specifically, three core optimization frameworks are formulated to maximize communication security while ensuring compliance with sensing-resistance (SR) constraints, tailored to varying levels of channel state information (CSI) availability. To solve the resulting non-convex problems, generalized Rayleigh quotient and generalized singular value decomposition are leveraged to reveal tractable problem structures and decouple the interaction between communication signals and artificial noise. Simulation results validate the proposed framework and demonstrate an explicit tradeoff between secrecy enhancement and localization resilience.
To address the dual threats of eavesdropping and active detection posed by an illegal autonomous aerial vehicle (AAV), this paper proposes a novel reconfigurable intelligent surface (RIS)-assisted integrated sensing and communication (ISAC) secure cooperation scheme. By leveraging the wide-area coverage of the high-altitude platform (HAP) and the channel reconstruction of RIS, the proposed scheme enables covert downlink command transmission from the HAP to the ground gateway while ensuring secure uplink data transmission for ground users. Specifically, the HAP, equipped with both communication and sensing capabilities, interacts with the ground gateway with the assistance of RIS, while continuously sensing and estimating the positions of an AAV using reflected echoes. For the mobile AAV, a factor graph optimization (FGO) method is proposed to achieve accurate AAV state estimation by exploiting the temporal correlation of continuous observation data. Based on the estimated AAV coordinates, we obtain the channel state and further analyze the detection performance of AAV in coherent and non-coherent detection scenarios, deriving closed-form solutions for the detection error probability (DEP) and the optimal detection threshold. Building on the above analysis, an optimization problem is formulated to maximize the effective covert rate (ECR), subject to constraints of sensing accuracy and covertness. For this high-dimensional and non-convex problem, a channel-aware graph neural network (CA-GNN) algorithm is proposed to jointly optimize sensing and communication beamforming as well as RIS phase shifts. Simulation results demonstrate the superiority of the proposed scheme in improving system security and covert performance. Compared to the genetic algorithm-based scheme and the deep neural network (DNN) scheme, the proposed scheme improves the ECR by 34.5% and 21.5%, respectively.
High altitude platform stations (HAPSs) are becoming a key component of future non-terrestrial networks (NTNs). HAPSs can serve a larger area than uncrewed aerial vehicles (UAVs) and offer lower propagation latency, maintenance expense, and energy costs than satellites. A major application of HAPSs is to serve the areas where terrestrial network (TN) deployment is infeasible, especially in hard-to-reach areas and post-disaster areas. For instance, in the Amazon rainforest, the Mediterranean region, and deserts, TN deployment is severely constrained by geographical and environmental conditions. Only areas close to transportation networks or coastlines can be covered, while large areas remain uncovered. Such coverage holes in hard-to-reach areas are typically overlooked in existing literature. Motivated by these realistic cases, in this paper, we use tools from stochastic geometry to mathematically model hard-to-reach areas where cellular terrestrial infrastructure only exists at their perimeter. We propose to deploy a HAPS constellation over this hard-to-reach area to enhance connectivity. For that setup, we derive the downlink (DL) and uplink (UL) coverage performance of the considered user equipment (UE) as a function of the location of the UE inside the coverage hole. Our results show how the number of HAPSs, beamwidth, and HAPS altitude affect the DL and UL coverage probabilities. Finally, we provide multiple useful guidelines for future HAPS deployment.
Joint communication and radar sensing, also known as integrated sensing and communication (ISAC), is evolving current networks into multifunctional systems with significant potential. This work investigates the ISAC multiple access problem in the finite blocklength (FBL) regime, which has been largely overlooked in existing studies. A dual-functional ISAC receiver is considered, which simultaneously decodes information from multiple users and performs channel state sensing in uplink transmission. Overall, fundamental limits on energy efficiency and communication-sensing tradeoffs are analyzed for dual-functional ISAC multiple access systems. Specifically, achievability bounds on energy efficiency and the corresponding performance floors of decoding error and sensing accuracy are characterized under fixed blocklength and energy constraints. The FBL energy efficiency of representative ISAC multiple access schemes, including TDMA, ALOHA, and power domain and code domain nonorthogonal multiple access (PD/CD-NOMA), is compared under different user densities, demonstrating that CD-NOMA exhibits the most robust energy efficiency against access density, while the others perform favorably only at low densities. The tightness of the derived bounds is further examined and proved by relating to the derived exact pairwise error probability. Additionally, the communication-sensing tradeoffs for dual-functional ISAC multiple access in FBL regime are revealed. Moreover, we derive a universal Cramér–Rao bound (CRB) for generic sensing parameters to demonstrate the effectiveness of treating channel estimation as a high-level sensing objective, and further validate this approach through an illustrative example based on 3GPP-oriented channel modeling. Numerical results corroborate the energy efficiency analysis, bound tightness, and communication-sensing tradeoffs, providing insights for dualfunctional ISAC multiple access systems.
Joint beamforming optimization is a fundamental challenge in achieving efficient integrated sensing and communication (ISAC) for cell-free massive multiple-input multiple-output (MIMO) systems. In the beamforming-vector domain, the design problem is nonconvex because both the transmit-side sensing objective and the user signal-to-interference-plus-noise ratio (SINR) constraints depend quadratically on the transmit beamforming vectors. A common approach is semidefinite relaxation (SDR), where the rank-one constraints on the transmit covariance matrices are removed to obtain a convex semidefinite programming (SDP) problem. However, the relaxed SDP solution may have rank greater than one and therefore cannot be directly mapped to feasible transmit beamforming vectors. Consequently, Gaussian randomization is required to recover rank-one beamforming candidates. In contrast to the SDR approach, this paper proposes a Difference-of-Convex Algorithm (DCA)-based optimization framework that directly enforces the rank-one structure during the optimization process. Specifically, the rank-one requirement is represented by a DC penalty term on each transmit covariance matrix, which regularizes the iterates toward rank-one beamforming covariance matrices. Based on this framework, we develop a DCA implementation tailored to coordinated cell-free architectures, incorporating efficient subgradient computation and an adaptive penalty scheduling mechanism. Simulation results show that the proposed mechanism achieves similar sensing SNR and minimum communication SINR to the Joint Sensing and Communication optimization based on the Semidefinite Programming (JSC-SDP) mechanism. Furthermore, the proposed DCA reduces the rank-1 violation rate by up to 81% and reduces the normalized communication-sensing beam overlap of approximately 33.3%.
In this paper, we investigate the transmit beamforming design for a general multiple-input multiple-output integrated sensing and communication (MIMO-ISAC) system, where a dual-functional base station (BS) simultaneously communicates with multiple multi-antenna communication users (CUs) and detects the multiple targets in the presence of multiple signal-dependent clutters. In multi-target sensing scenarios, it is essential to ensure a baseline sensing performance for all targets, since the missed detection of even a single target may degrade the reliability of the overall system. To this end, we adopt the max–min fairness (MMF) criterion and formulate a minimum sensing mutual information (SMI) maximization problem under communication quality-of-service (QoS) requirements and a total transmit power constraint. Unlike conventional CVX-based methods, we develop a novel low-complexity algorithm that integrates epigraph reformulation, matrix fractional programming (FP), and the projected extragradient (PEG) method to efficiently handle the formulated non-convex and non-smooth optimization problem. In addition, two benchmark algorithms are introduced to verify the effectiveness of the proposed MMF-oriented design. Furthermore, the proposed algorithm is extended to scenarios in which the BS has imperfect channel state information (CSI) for both the communication and sensing channels. Finally, numerical simulation results demonstrate the superiority of the proposed design under both perfect and imperfect CSI conditions. In particular, the proposed schemes achieve performance comparable to that of conventional CVX-based approaches while reducing the computational time by approximately 70%.
Integrated sensing, communication, and computation (ISCC) networks have emerged as a key enabler for intelligent applications. Most existing works focus on small-scale ISCC systems with a limited number of nodes, making it difficult to provide guidance for large-scale ISCC network design. This paper bridges this gap by developing an analytical framework for large-scale ISCC networks with massive spatially random distributed nodes, in which both periodic requests and aperiodic requests are considered. To support dynamic sensing requests, integrated sensing and communication (ISAC) devices perform wireless sensing to collect sensing data, which will be processed through either edge offloading or local computing. Given the model, the queue dynamics and interference intensity are first analyzed for interference characterization. Next, the b-th moments and the meta distribution of successful sensing probability and successful communication probability are derived. Based on these results, the sensing delay, communication delay, and computing delay are jointly analyzed to evaluate the end to end (E2E) delay performance. Simulation results validate the accuracy of the proposed analytical framework in terms of ISAC service reliability and E2E delay. It shows that compared with the benchmark analytical frameworks that underestimate delay, the proposed framework provides accurate performance characterization to avoid unreliable network deployment. Further, the impact of network parameters on delay performance and the tradeoffs between sensing reliability and E2E delay, communication reliability and E2E delay, as well as local and edge computing delays, is numerically analyzed. These results can provide useful insights to enable appropriate system design based on multiple performance metrics.
Fluid antenna systems (FAS) have emerged as a promising paradigm for wireless communications, enabling channel reconfigurability that offers a novel spatial degree of freedom. Nevertheless, efficiently acquiring accurate and high-resolution channel state information (CSI) in FAS remains challenging, primarily due to its dynamic spatial structure and limited coherence time. This paper proposes a novel diffusion framework that takes the partially observed CSI matrix as the terminal state of the Markov chain and operates exclusively on the unobserved ports. Built upon this framework, we design a UNet-based architecture, termed the channel extrapolation UNet (CEUNet), that integrates modified MaxViT (mMaxViT) blocks and residual blocks (ResBlocks) to jointly capture local and global channel dependencies for accurate CSI extrapolation. Extensive experiments on the Jakes’ channel model are conducted to evaluate CEUNet. Numerical results show that CEUNet consistently outperforms state-of-the-art deep learning models in estimation accuracy across all signal-to-noise ratios and observation ratios, even with only two sampling steps. Furthermore, a comprehensive complexity analysis is conducted to compare the computational efficiency of CEUNet with that of the baseline models, while ablation studies are carried out to quantitatively evaluate the contribution of each component integrated into the proposed CEUNet.
A pinching antenna systems (PASS)-aided over-the-air computation (OAC) framework is proposed, where edge servers (ESs) deliver an aggregate to the edge device (ED). To reduce the aggregation distortion, a unified mean-square error (MSE) objective is defined over the ES pinching beamformers, ED combiner, and effective channel that incorporates in-waveguide propagation and programmable radiation. By exploiting different configurations of pinching antennas (PAs), two aggregation protocols are proposed, i.e., in-waveguide aggregation (IWA) and free-space aggregation (FSA). In IWA, aggregation is formed within an interconnected waveguide grid and is forwarded through a single active PA. A rank-one effective channel is thus induced, which enables elimination of the continuous transceiver beam variables. An elimination-based branch-and-bound (EBnB) algorithm is developed to attain global optimality over grid-based PA locations and transceiver beams. In FSA, aggregation is formed in free space through multi-point radiation over isolated waveguides. A hierarchical algorithm termed GPPO is developed, where discrete activations and continuous placements of PAs are produced by a geometry-aware graph neural network policy trained via proximal policy optimization, while an inner solver alternately updates the transceiver beams in closed form. Simulation results validate that i) IWA generally provides lower MSE and more robust performance due to in-waveguide aggregation and single-PA radiation, at the cost of higher deployment-control complexity; ii) FSA can achieve lower MSE under favorable geometries with sufficient spatial degrees of freedom, but is more sensitive to the deployment geometry; iii) PASS-aided OAC achieves lower MSE than conventional multi-input multi-output (MIMO); and iv) the proposed EBnB and GPPO algorithms outperform representative baselines in terms of MSE.
This paper investigates a secure massive multiple-input multiple-output (mMIMO) integrated sensing and communication (ISAC) system and proposes a framework capable of precisely localizing an active eavesdropper (Eve) while simultaneously ensuring secure communication for legitimate user equipment (UEs). Compared with existing ISAC studies that primarily focus on sensing in the downlink phase, the proposed design leverages the uplink and downlink structure of time-division duplexing (TDD)-based ISAC systems by explicitly exploiting uplink training information to assist downlink localization. In the uplink, pilot-based channel estimation is performed in the presence of a pilot spoofing attack (PSA), enabling reliable detection of Eve’s angle of arrival (AoA). This coarse AoA estimate is then used in the downlink to form a dedicated sensing region that achieves precise, multipath-resilient direct localization of the Eve. Closed0form expressions for the signal-to-interference-plus-noise ratios (SINRs) at the legitimate UEs and Eve are derived, facilitating a theoretical evaluation of the system’s secrecy spectral efficiency (SSE). A tractable closed-form expression for the position error bound (PEB) is also derived. To jointly optimize sensing and communication performance, we propose a power allocation problem that maximizes the SSE while satisfying a given mainlobe-to-average-sidelobe ratio (MASR) threshold and power constraints. Numerical results demonstrate that the proposed joint uplink and downlink ISAC framework achieves substantially improved localization accuracy in multipath scenarios and significantly enhances secrecy performance, compared to conventional secure ISAC systems.
Multiple access (MA) design is investigated to facilitate pinching-antenna systems (PASS)-based multi-user communications. By exploiting the newly introduced waveguide domain and existing frequency domain, two MA schemes are explored, namely pure waveguide division multiple access (WDMA) and hybrid WDMA. For each MA scheme, the corresponding resource allocation problem is formulated to maximize the rate fairness via the joint optimization of pinching beamforming and power allocation. For both schemes, a majorization-minimization (MM)-based alternating optimization (AO) algorithm is proposed that alternately optimizes pinching beamforming and transmit power. A low-complexity framework is further developed, including a two-stage pinching beamforming design and successive convex approximation (SCA)-based power allocation. Numerical results demonstrate that: 1) PASS significantly improve communication rate performance over conventional antenna systems; 2) The proposed MM-based AO algorithm provides higher performance at the cost of increased complexity, while the low-complexity framework achieves comparable performance with lower computational complexity; 3) Pure WDMA achieves better performance compared to hybrid WDMA, efficiently supporting multi-user communications enabled by pinching beamforming.
Low Earth Orbit (LEO) satellite networks are envisioned to provide wide-area connectivity for a massive number of user terminals. Realizing this vision requires efficient resource allocation, including user grouping, beamforming (BF) and power control, that adapts to the uneven spatio-temporal distribution of user location and traffic demand. Existing allocation schemes primarily rely on channel state information (CSI) while neglecting the stochastic nature of traffic arrivals, leading to spectrum efficiency degradation and reduced system stability. To solve this problem, we propose a two-timescale joint optimization framework that integrates CSI and queue state information (QSI) to enable efficient resource allocation for LEO satellite systems. On the slow timescale, we propose a novel two-stage user grouping scheme that combines traffic-aware K-means++ clustering with a grid-based spatial dividing method to construct user groups with traffic-aware load characteristics and improved spatial orthogonality. These pre-constructed groups serve as scheduling candidates for the fast timescale, enabling structured decisions with reduced complexity and improved spatial separability among co-scheduled users. On the fast timescale, we develop a queue-aware group scheduling, BF, and power control algorithm based on the CSI and QSI, aiming to maximize queue-weighted utility under signal-to-interference-plus-noise ratio and power constraints. Finally, simulation results verify that the proposed approach significantly improves system service performance while ensuring queue stability, outperforming existing benchmark methods.
