With the development of the Starlink network, satellite communication has become a primary means of Internet access in remote areas. Leveraging the ubiquitous downlink from Low-Earth-Orbit (LEO) satellites and the portability of ground stations, this paper explores the feasibility of using Starlink downlink signals as an illuminating signal for Ambient Backscatter Communication (AmBC). We propose an AmBC system that operates in the Ku-band, utilizing Starlink downlink signals. Furthermore, we analyze the impact factors for AmBC performance and derive its analytical expression. Numerical simulation results demonstrate that the AmBC based on the Starlink downlink achieves a Bit Error Rate (BER) on the order of 10(-2), highlighting its potential for practical applications.
As a pivotal enabling technology for 6G wireless networks, extra-large scale multiple-input multiple-output (XL-MIMO) systems present unique hybrid-field channel character-istics that necessitate joint consideration of near-field and far-field components. While existing solutions predominantly focus on pure near-field scenarios, the critical challenge in hybrid-field environments lies in effectively decoupling these spatially diverse channel components under practical constraints. This paper proposes a novel low overhead channel estimation frame-work for hybrid-field XL-MIMO systems. Our approach begins by reconstructing the angular domain channel representation through optimized hybrid beamforming architecture. Specifically, we develop a phase shift matrix optimization strategy that enables accurate angular domain recovery under stringent pilot constraints. The proposed spatial filtering technique then achieves effective field separation: far-field components are extracted through joint analysis of beamwidth characteristics and amplitude distributions, while residual near-field components are sub-sequently recovered via compressed sensing-based reconstruction. The proposed algorithm can reconstruct hybrid-field channel estimation with low pilot overhead.
This paper investigates content uploading in spatialaerial integrated low altitude networks (SALAN), where the distributed low altitude terminals (LATs) are mobile and far away from terrestrial communication infrastructures. In the SALAN, LATs can upload their data to a satellite access point through multiple flexibly deployable aerial relays (ARs). To minimize the maximum uploading latency of LATs, we formulate the AR trajectory planning and resource allocation (ARTPRA) problem and solve it by decomposing it into two sub-problems, i.e., an upload rate control (URC) sub-problem, and an AR trajectory planning and frequency band management (ARTPFM) sub-problem. We prove that the optimal solution to the URC sub-problem can be achieved by solving a series of linear programming problems and propose a multi-agent deep reinforcement learning (DRL)-based framework to solve the ARTPFM sub-problem. Extensive experimental results show that our approach achieves very low offline training complexity for the ARTPFM framework and outperforms the benchmark approaches in terms of latency for uploading tasks, empirical success rate, and online computational time for content uploading.
Remote inference, where sensors acquire data and transmit compressed features to a remote server for inference, plays a key role in numerous applications. A common remote inference task is direction of arrival (DoA) estimation, where the sensed data is used for localizing multiple sources. Traditional methods require the sensor to downstream raw wideband data, leading to increased latency and spectral inefficiency. In this work, we propose Remote SubspaceNet, a deep neural network (DNN)-aided remote inference framework that enables interpretable, low-latency, and progressively refined DoA estimation at the server. Remote SubspaceNet integrates DNN-based feature extraction with learned vector quantization and subspace-based inference techniques, learning a single quantization codebook that supports successively refined DoA recovery. We demonstrate that Remote SubspaceNet accurately estimates DoAs across varying bit budgets, significantly reducing communication latency while maintaining interpretability and robustness.
In this paper, we investigate the problem of spatial non-stationary channel estimation for near-field extremely large-scale multiple-input multiple-output (XL-MIMO) systems. Specifically, we propose a mixed model- and data-driven deep learning (DL) approach that achieves accurate near-field channel estimation under spatial non-stationary effects while maintaining low computational complexity. Our approach first exploits the polar-domain sparsity of the near-field channel to develop an unfolding network for efficient on-grid sparse channel recovery. This network is designed by unfolding the inverse-free variational Bayesian inference (IF-VBI) algorithm, referred to as VBI-Net. Moreover, to mitigate the energy leakage effect inherent in ongrid methods and the estimation error introduced by spatial non-stationary effects, we further enhance channel estimation accuracy by utilizing a data-driven deep learning framework. Simulation results validate that our proposed mixed model- and data-driven DL approach significantly outperforms existing state-of-the-art schemes, validating its effectiveness and superiority.
The spectrum is a shared, congested, and densely used resource. The emerging sixth-generation wireless (6 G) systems intend to alleviate this problem by integrating radio frequency sensing and wireless communications functionalities, enabling spectrum, hardware, and antenna resource sharing. Agile spectrum use requires situational awareness (SA) about the state of the radio environment. Automatic waveform recognition (AWR) is an important part of building and maintaining SA needed in cognitive radios, radars, and signal intelligence. Commonly, a convolutional neural network (CNN) has been paired with time-frequency (TF) imaging to distinguish among different low probability of intercept/detection (LPI/LPD) waveforms. In this paper, we thoroughly investigate the effect of the TF resolution in the classification accuracy of twelve LPI/LPD radar waveforms using different variants of the Fourier-based synchrosqueezing transform (FSST) to generate time-frequency images (TFIs). The pros and cons of downsampling in time and frequency domains are analyzed at the Heisenberg-Gabor limit. Practical classification results are presented at different sensing channel conditions.
This paper introduces a novel quantum-assisted algorithm for partial sorting, which is integrated with a decoding algorithm for spatially modulated wireless signals. The decoder achieves near-optimal bit error rate performance, closely matching that of the maximum-likelihood decoder, while significantly reducing computational complexity. Additionally, integrating the quantum partial sorting algorithm lowers the query complexity. Although quantum algorithms introduce a small failure probability, resulting in an error floor, simulations validate the scalability and practical efficiency of the utilized decoder with quantumassisted sorting.
We consider the problem of integrated sensing and communication (ISAC) involving a massive number of users in an unsourced random access (URA) system, called unsourced ISAC (UNISAC). In UNISAC, there are two sets of communication and sensing users that transmit signals in a random-access way. Hence, the received signal of each user is heavily corrupted by interference from numerous other users, making it challenging to extract individual transmissions. UNISAC is designed to decode the message sequences from communication users while simultaneously detecting active sensing users and estimating their angles of arrivals, regardless of the senders' identities. We propose a practical model for UNISAC that demonstrates the feasibility of the approach, achieving performance comparable to-or even surpassing-known ISAC results in certain regimes while being lower in others.
The Digital Enhanced Cordless Telecommunications (DECT-2020) New Radio (NR+), standardized by the European Telecommunications Standards Institute (ETSI), addresses International Telecommunication Union (ITU) IMT-2020 requirements for Internet of Things (IoT) and Industrial IoT. Although DECT-2020 NR+ offers robust connectivity, it does not support ultra-low power or batteryless operation. The 3rd Generation Partnership Project's (3GPP's) Ambient Internet of Things (AIoT) initiative targets ultra-low power IoT devices operating on harvested ambient energy, with backscatter communication as a promising technology. This paper integrates backscatter-based AIoT functionality into DECT networks, leveraging existing DECT transmissions and data-assisted channel estimation to recover backscattered information. An Ambient Backscatter Communication (AmBC) receiver is symbiotically embedded within the DECT receiver, using Received Signal Strength (RSS) of DECT data traffic. We open-sourced the Matlab code for the DECT receiver in a Software-Defined Radio (SDR) platform. Over-the-air experiments demonstrate feasibility, achieving Bit Error Rate (BER) performance as low as 0.02 under optimal conditions.
In this contribution, we investigate the identification of orthogonal frequency division multiplexing (OFDM) and OFDM with index modulation (OFDM-IM) signals under various practical constraints. Firstly, the utilization of maximum likelihood estimation (MLE) is considered, while a computationally efficient statistical method based on the variance of received signals is also presented, aiming at significant complexity reduction. Subsequently, a robust convolutional neural network (CNN)-based method is conceived to overcome practical limitations stemming from insufficient prior knowledge, thereby facilitating reliable signal identification. Finally, simulation results demonstrate that the developed variance-based method can overcome the high-complexity challenge of its MLE-based counterpart, while the proposed CNN-based method maintains consistent and robust identification accuracy, highlighting its suitability for practical applications where channel conditions and noise statistics are uncertain.
This paper addresses the power control design for a cell-free massive MIMO (CF-mMIMO) system that performs integrated sensing and communications (ISAC). Specifically, the case where many access points are deployed to simultaneously communicate with mobile users and monitor the surrounding environment at the same time-frequency slot is considered. On top of the user-centric architecture used for the data services, a target-centric approach is introduced for the detection tasks. As a valuable performance metric, we derive the receive sensing signal-to-noise (SNR) ratio under generalized likelihood ratio test processing. Based on that, we formulate a quality-of-service (QoS) scheme that maximizes the two figures of merit: achievable data rate and effective sensing SNR. Simulations demonstrate that our proposal surpasses orthogonal resource algorithms, underscoring the potential of ISAC-enabled CF-mMIMO networks.
Radar systems' angular resolution depends on the available aperture size, which, for a practically limited aperture, impedes radar's ability to effectively resolve closely-spaced targets. A conventional approach to improving angular resolution would be to increase the number of antennas in a larger aperture, but this introduces power consumption, computational complexity, and area occupancy costs. A standard approach to reducing the number of antennas would be to sparsify the antenna array while maintaining the same aperture. These types of arrays are known as thinned arrays. Thinned arrays, however, have difficulty in resolving closely-spaced targets due to the manifestation of sidelobes causing the degradation of antenna directivity and/or gain. In this paper, we propose convolving two thinned arrays to attenuate sidelobes and improve closely-spaced target separation. The convolution of the two thinned arrays is analyzed and used to define atoms of a dictionary of a "virtual array", allowing localization using the orthogonal matching pursuit (OMP) algorithm. In the simulated results, the proposed method is compared to two methods, one thinned array with twice the number of antennas as one of the thinned arrays used in convolution, and the average angle of arrival (AoA) estimates of each thinned array used in convolution. The results demonstrate the superiority of our proposed method in terms of improving closely-spaced target separation, even in the presence of an increased number of
Due to the reduced level of interference, mmWave networks favour spectrum sharing among operators. To further enhance spectrum utilization and licensing flexibility, spectrum owners can share their licenses with secondary users under some transmission restrictions, resulting in a cognitive mmWave network. Given the high directionality and sensitivity of mmWave communication towards blockages, spectrum usage opportunities can be significantly improved by incorporating directional gain and blockage state in these restrictions. Utilizing the tools from stochastic geometry, we develop an analytical framework to study the impact of blockages on the performance of such mmWave cognitive networks. We characterize the secondary network's transmission opportunities and the coverage probability of the primary and secondary users. We show that blockages can improve the secondary activity and performance of both types of links by reducing the overall interference.
Recovering environmental information from the wireless channel information is one of the dual goals of integrated sensing and communication (ISAC). However, traditional methods of environment sensing fail to fully exploit the physical information of wireless signal propagation, which limits the sensing accuracy and efficiency. In this paper, we consider a multi-view environment sensing scenario, and formulate the problem of reconstructing the unknown scatterer as optimizing a neural surface. We first model the propagation mechanism of wireless signals based on the theory of electromagnetic scattering and ray tracing, and then parametrize the scatterer surface with a neural network that represents the signed distance function (SDF). After modeling the volume occupancy probabilities in the scene, we apply ray tracing to predict the channel response, which is differentiable w.r.t the network parameters. Therefore, we can optimize the neural network by minimizing the prediction error to reconstruct the scatterer shape. Our method combines electromagnetic wave theory with neural network optimization, and exploits multi-view channel information to achieve high reconstruction quality. Simulation results have validated the effectiveness of the proposed reconstruction method.
Vehicular platooning refers to a group of vehicular users (VUs) traveling together along a common route segment, offering strategic benefits such as reduced fuel costs, lower emissions, and improved traffic flow. Highways offer a natural setting for platooning due to extended travel distances, yet their potential remains underexplored, particularly in terms of communication and connectivity. Given that effective platooning relies on robust vehicle-to-vehicle (V2V) and vehicle-to-infrastructure (V2I) links, understanding connectivity dynamics on highways is essential. In this paper, we model and analyze the spatial distribution and network load characteristics of platooned and non-platooned vehicular traffic on a highway. Roadside units (RSUs) are distributed according to a homogeneous one-dimensional (1D) Poisson point process (PPP), while VUs are modeled using a 1D PPP for non-platooned traffic and a 1D Matern cluster process (MCP) for platoons. We investigate the resulting RSU load distribution, representing the number of VUs served under varying traffic densities and platooning scenarios.
Rate-splitting multiple access (RSMA) has emerged as a promising non-orthogonal transmission scheme with efficient interference management. On the other hand, a hybrid active-passive reconfigurable intelligent surface (HRIS) can enhance wireless channels by simultaneously reflecting and amplifying incident signals. In this paper, we explore the synergistic advantages of integrating these technologies within an HRIS-assisted RSMA system. We derive highly accurate expressions for the system's outage probability and ergodic capacity under Nakagami-$m$ fading channels and successive interference cancellation errors, leveraging the moment-matching technique. Furthermore, we present the asymptotic expressions for the outage probability and ergodic capacity to unveil important insights into the diversity order and coding gains of the system. Finally, we provide extensive numerical results to validate the analytical findings.
This work addresses the problem of location and velocity estimation for a moving source in the near-field region of an extremely large antenna array (ELAA). The velocity of the source moving within the near-field region of an antenna array is influenced by both radial and transverse velocity components. Therefore, accurately estimating the location and velocity of a near-field source requires determining four unknowns: direction of arrival (DOA), range, radial velocity, and transverse velocity. One approach that can be used to tackle this problem is the maximum likelihood (ML) method. However, the conventional ML technique is computationally infeasible due to its high complexity. To address this issue, we first employ an ML method to estimate the DOA and radial velocity using a sub-array of an ELAA. Then, using another ML method and utilizing the entire ELAA along with the estimated DOA and radial velocity, we estimate the range and transverse velocity. Simulation results confirm the effectiveness of the proposed method.
We provide wireless network virtualization (WNV) for multiple service providers (SPs) that are assumed to virtually serve multi-antenna users via a base station (BS) with multiple antennas. The virtualization of this multi-user multiple-input multiple-output (MU-MIMO) system is managed by an infrastructure provider (InP) that owns the communication equipment. By exploiting the antennas at both the BS and the user devices, we jointly design the uplink receive beamforming at the BS and the transmit beamforming at the user devices, achieving service isolation at the physical layer. This WNV is formulated as a non-convex optimization problem in the transmit and receive beamforming vectors. We derive separately optimal closed-form and semi-closed form solutions for the beamforming vectors, and use these solutions in alternating iterations to solve the original problem. Our simulation results show that the proposed solution enables effective network virtualization, to support the independent operation of multiple SPs, while retaining efficiency no less than non-virtualized operation, and it substantially outperforms traditional WNV based on strict resource separation.
In millimeter-wave (mmW) networks, large antenna arrays can be deployed to combat high signal attenuation, yet creating also near-field (NF) effects in the relative proximity of the antenna system. While utilizing frequency-selective rainbow beams enabled by true-time-delay (TTD) analog beamformer, we study the mmW network localization capabilities in the NF domain via deep learning neural networks. By leveraging the unique properties of the rainbow beams, we show that the proposed deep learning model, referred to as RaiNet, is capable of accurately positioning the user using a single channel response measurement. The provided numerical results at different carrier frequencies show that the proposed deep learning approach enables significant improvements in localization accuracy, compared to the state-of-the-art benchmark methods. The study thus paves the way for advanced localization techniques in 6G systems, contributing to the development of more efficient and intelligent future networks.
We consider the uplink of massive MIMO systems with one-bit quantization and propose an iterative detection and decoding (IDD) receiver based on a factor graph representation and the sum-product algorithm. We derive the relevant sum-product messages, introduce Gaussian approximations to control computational complexity, and describe the message schedule. Numerical experiments demonstrate that the BLER performance of our receiver is comparable to the state of the art, even though our method has much lower complexity - quadratic rather than exponential in the number of users - and hence scales well to large system configurations. We show that a proper scheduling of the decoding iterations substantially improves the error rate performance without increasing the computational complexity. In particular, we find that it is not beneficial to schedule the majority of decoding iterations in the initial detection iterations.
