Precise distance measurement using wireless signals is essential for applications such as structural health monitoring, autonomous navigation, and asset tracking. Traditional high-accuracy ranging techniques typically rely on large signal bandwidths, often several hundred megahertz, to resolve carrier-phase ambiguities and achieve centimeter-level precision. However, this results in increased system complexity and high spectral resource consumption. To address these challenges, this paper proposes an optimized dual-differential carrier phase measurement method that achieves centimeter-level accuracy using only a few of discrete frequency bands, each occupying a bandwidth of approximately 20 MHz. Carrier-phase measurements at the selected frequencies are first obtained and wrapped within the (−π,+π) range. Dual-differential operations are then applied between adjacent positions to eliminate systematic phase biases and improve estimation robustness. A weighted least-squares estimator provides initial float ambiguity estimates, which are subsequently resolved using the Modified Least-squares Ambiguity decorrelation adjustment (MLAMBDA) algorithm. Compared with conventional ranging methods, the proposed approach significantly reduces bandwidth requirements while maintaining sub-5 cm relative positioning accuracy. It is particularly well-suited for deployment in bandwidth-constrained Integrated Sensing and Communication (ISAC) applications.

Sensing-assisted communication plays a critical role in integrated sensing and communication (ISAC) systems, where environmental awareness can significantly enhance communication performance. In this paper, we propose a novel indoor environment reconstruction method for Sub-6GHz systems. The approach first leverages multiband channel state information (CSI) to enable accurate localization of user equipment (UE). Based on the estimated UE positions and multiband CSI, we further estimate the distances of multipath components (MPCs). These distances, combined with UE positions, are used to infer the positions of virtual-access points (VAPs), which serve as key intermediaries for identifying environmental reflectors. By aggregating information across multiple UEs, we compute a reflector point cloud that supports high-fidelity reconstruction of the indoor environment. Our proposed algorithm is validated through simulations, achieving a corner point localization error of 0.03 meters and an intersection over union (IoU) of 0.98. Experimental results demonstrate the effectiveness of the method in reconstructing typical rectangular indoor layouts using only Sub-6GHz CSI.

Faced with the increasing demand for indoor localization, global navigation satellite systems perform significantly worse in indoor environments than in outdoor scenarios. Existing indoor algorithms also have high requirements for the environment and terminal devices. To address this, we propose a new simultaneous localization and mapping algorithm that improves the Rao-Blackwellized probability hypothesis density filter, achieving target tracking using only Time of Flight with the assistance of an inaccurate floor plan and we incorporate a new data association method into this framework. Additionally, given that most current tracking algorithms are based on a known initial position, we have proposed an algorithm to determine the target's initial position. Simulation results demonstrate the effectiveness of the algorithm.

The vehicle key tracking technology plays a crucial role in modern vehicles. It not only enhances the security of the vehicle but also provides a more convenient user experience, making it an indispensable component of modern smart vehicles. This paper presents a vehicle key tracking system based on the identification of non-line of sight (NLOS) anchors, enabling more accurate tracking. The system first utilizes a direct positioning estimation (DPE) algorithm to estimate the target's position, followed by calculating the distances between the estimated position and anchors. These distances are then compared with the actual measurement data to determine an appropriate threshold for distinguishing between NLOS and LOS signals, with the ranging results from LOS signals used for tracking. Finally, the system was tested using a software-defined radio (SDR) platform. Experimental results demonstrate that the proposed algorithm achieves a median tracking error of 0.24 meters, compared to a median tracking error of 12 meters when tracking without using the NLOS identification algorithm. Furthermore, the 90 % tracking error for the circular trajectory is 0.72 meters.

Speaker embedding based zero-shot Text-to-Speech (TTS) systems enable high-quality speech synthesis for unseen speakers using minimal data. However, these systems are vulnerable to adversarial attacks, where an attacker introduces imperceptible perturbations to the original speaker’s audio waveform, leading to synthesized speech sounds like another person. This vulnerability poses significant security risks, including speaker identity spoofing and unauthorized voice manipulation. This paper investigates two primary defense strategies to address these threats: adversarial training and adversarial purification. Adversarial training enhances the model’s robustness by integrating adversarial examples during the training process, thereby improving resistance to such attacks. Adversarial purification, on the other hand, employs diffusion probabilistic models to revert adversarially perturbed audio to its clean form. Experimental results demonstrate that these defense mechanisms can significantly reduce the impact of adversarial perturbations, enhancing the security and reliability of speaker embedding based zero-shot TTS systems in adversarial environments.

The Global Navigation Satellite System (GNSS) offers advantages such as all-weather operability and extensive spatial coverage. Utilizing GNSS-reflected signals for ground synthetic aperture radar (SAR) imaging presents a cost-effective and widely applicable technical solution. However, the small bandwidth of GNSS signals results in inadequate resolution, posing challenges for practical applications. To address this issue, an SAR fusion imaging system model is established, consisting of multiple satellites and a single ground-fixed GNSS receiver. The relationship between the ambiguity function of GNSS signals and Fisher information is investigated, allowing for the derivation of the Cramer-Rao lower bound (CRLB) for the system, which is primarily influenced by the geometrical configuration of the bistatic setup. Subsequently, the CRLB expression is employed to identify the optimal resolution direction of the satellites for ground targets, and a dual-satellite SAR imaging fusion method based on optimal matching is proposed. The effectiveness of this approach is validated through simulations and real experimental data, demonstrating that the theoretically optimal resolution direction predicted by the CRLB aligns with the actual imaging results. Furthermore, the proposed method achieves higher resolution compared to traditional techniques, with the fused imaging results demonstrating a clear correspondence with the satellite imagery of the scene map.

In response to the drawbacks of indoor multi-station joint localization, this paper proposes a high-precision indoor single-station localization technique based on the multipath shape factor (SF). A geometric localization model was constructed using the reflection and direct path. The localization equation was constructed using time of flight (ToF) and angle of arrival (AoA), and the Cramér-Rao lower bound (CRLB) of localization error was derived. Through CRLB, this paper introduced multipath SFs and derived the calculation expression of SFs. However, localization accuracy is affected by observation errors and other factors, and no effective method has been proposed yet. To improve localization accuracy and reduce algorithm complexity, this paper first proposes a compressive sensing multidimensional parameter estimation algorithm based on particle swarm optimization (PSO). Then, this paper proposes a multipath selection algorithm based on the SF. Finally, we conduct the simulation using commercial-grade wireless ray propagation modeling software. The experimental results show that the median location error is around 0.18 m by combining the two algorithms proposed in this paper.

Ranging and localization are crucial elements in the field of sensing applications, with range-based localization techniques being a prevalent approach. However, traditional range-based localization methods are impacted by ranging accuracy, furthermore, ranging accuracy is related to bandwidth. In this paper, we achieve high resolution by splicing the channel state information (CSI) of multiple non-adjacent frequency bands. However, the multiband CSI of GHz-level sub-band spacing leads to ambiguous delay estimation. To address this issue, we first explain the reason for the delay ambiguity caused by multiband CSI. Then, a set of candidate delays is constructed using the multiband CSI estimation delay. We propose a lo-calization algorithm that utilizes the set of candidate delays to achieve accurate localization. In turn, we use the estimated accurate location for accurate ranging. Finally, we validate our proposed localization and ranging algorithm through simulation experiments, achieving a localization error of 0.02m in 90% of cases, and a ranging accuracy of 0.01m in 94% of cases. Experimental results demonstrate that the algorithm proposed in this paper effectively exploits the advantages of multiband CSI for enhanced ranging and localization.

As the demand for location-based services has surged, indoor localization has become a research hotspot. In this paper, we employ convolutional neural network (CNN) and configurable multipath fingerprint methods for indoor lo-calization. To validate this approach, we design an indoor localization system based on only a single access point (AP). In this system, we first configure active reflectors within the indoor environment to stabilize the configurable multipaths, and model the wireless channel under conditions where configurable multipaths exist. Next, we utilize the two-dimensional recursive applied and projected multiple signal classification algorithm to jointly estimate the angle of arrival and time of flight, and then convert the results into spectral images. Then, these spectral images are fed into our designed CNN model for training, and the model’s output is subsequently matched with the target location. Finally, we conduct experiments using the commercial-grade electromagnetic propagation simulation software Wireless InSite. The experimental results show that with two deployed reflectors, the median error is 0.40 m, which outperforms other localization systems based on a single AP. Moreover, the introduction of active reflectors further enhances the system’s generalization ability.

Utilizing the pseudo-labeling algorithm with large-scale unlabeled data becomes crucial for semi-supervised domain adaptation in speaker verification tasks. In this paper, we propose a novel pseudo-labeling method named Multi-objective Progressive Clustering (MoPC), specifically designed for semi-supervised domain adaptation. Firstly, we utilize limited labeled data from the target domain to derive domain-specific descriptors based on multiple distinct objectives, namely within-graph denoising, intra-class denoising and inter-class denoising. Then, the Infomap algorithm is adopted for embedding clustering, and the descriptors are leveraged to further refine the target domain’s pseudo-labels. Moreover, to further improve the quality of pseudo labels, we introduce the subcenter-purification and progressive-merging strategy for label denoising. Our proposed MoPC method achieves 4.95% EER and ranked the 1stplace on the evaluation set of VoxSRC 2023 track 3. We also conduct additional experiments on the FFSVC dataset and yield promising results.

Voice conversion (VC) systems can transform audio to mimic another speaker’s voice, thereby attacking speaker verification (SV) systems. However, ongoing studies on source speaker verification (SSV) are hindered by limited data availability and methodological constraints. This paper presents the Source Speaker Tracking Challenge (SSTC) on STL 2024, which aims to fill the gap in the database and benchmark for the SSV task. In this study, we generate a large-scale converted speech database with 16 common VC methods and train a batch of baseline systems based on the MFA-Conformer architecture. In addition, we introduced a related task called conversion method recognition, with the aim of assisting the SSV task. We expect SSTC to be a platform for advancing the development of the SSV task and provide further insights into the performance and limitations of current SV systems against VC attacks. Further details about SSTC can be found here1.1https://sstc-challenge.github.io/

Indoor accurate ranging using WiFi signals is a challenge since it is affected by the signal bandwidth, indoor environment, and phase distortions introduced by the underlying hardware. In this paper, we improve channel impulse response (CIR) resolution by splicing channel state information (CSI) measurements from multiple non-contiguous bands. However, most previous efforts have directly used non-contiguous CSI measurements for parameter estimation, which leads to multi-band gain loss and ranging performance degradation. To solve this problem, we propose a novel time of flight (TOF) estimation scheme that includes two stages. In the first stage, we build an equivalent optimization model of the inverse non-uniform discrete Fourier transform (INDFT) to transform the CSI samples to the CIR. In the second stage, we construct a time domain super-resolution TOF estimation model and combine the estimation results of the first stage to obtain the distance information using a subspace projection method. Finally, we conduct the simulation using commercial-grade wireless ray propagation modeling software. Based on the simulation result, the ranging error of our approach is less than 10 cm 85 % with the spliced bandwidth of 240 MHz.

Indoor localization systems in wireless sensor networks (WSNs), such as WiFi BLE, are still not widely adopted despite years of continuous research. One possible impediment is that many prior systems perform worse in non-line-of-sight (NLoS) indoor scenarios than in line-of-sight (LoS) scenarios or even fail to work. Nevertheless, realizing localization in both LoS and NLoS scenarios is indispensable for a viable system working in indoor environments. Since existing state-of-the-art solutions mainly aim to localize a target in LoS indoor scenarios, we present 3DLoc that uses multipath channel parameters to achieve 3D localization with a single AP in indoor NLoS scenarios. 3DLoc introduces deployable and simple custom-designed radio relays to assist localization, which can receive signals from the target and forward them to the AP in NLoS scenarios. Field trials conducted in a real NLoS scenario show that our system achieves a median 3D localization accuracy of 86 cm with a uniform circular array (UCA) in NLoS conditions.

Tracking is promising in various applications, such as elderly care and security monitoring. Prior works require multiple access points (AP) deployed in the tracking region or to utilize elaborately selected multipaths in an indoor environment to achieve tracking. However, this prevents their wide adoption because scenarios like homes typically only have a single AP, and using multipaths is vulnerable to unexpected interferences due to the complex indoor environment. This paper proposes a tracking system using only a single AP. The insight behind the proposed system is to mine the multi-dimensional channel parameters from the direct path to realize tracking. Therefore, we develop a model that establishes the correlation between the target motion and channel parameters, including the direct path's angle of arrival (AoA) and angle of departure (AoD). Then, we can formulate the tracking problem as the state transition with the observations. Since the nonlinear observation equation, we use particle filters to solve the problem accurately. Finally, we validate the proposed system via simulation using ray-tracing software. The simulation results show that the median tracking error is 0.21 m based on a 20 dB signal-to-noise ratio (SNR), comparable to previous state-of-the-art systems.

Indoor localization using WiFi signal has received great attentions since it is ubiquitous. So, in this paper we propose the design, implementation and evaluation of a real-time indoor localization system using commodity WiFi signal. The proposed system can localize the target by using Angle of Arrival (AOA) without any hardware modification and large localization latency. The contributions of this paper are follows. Firstly, an Angle estimation algorithm based on WiFi signals is proposed, which can quickly estimate AOA of Line of Sight (LOS) path in the case of fewer antennas and packets, ensuring real-time localization. Secondly, the influence of multipath signals on the energy spectrum of direct signals is analyzed by using the IEEE 802.11 Saleh-Valenzuela (S-V) channel model. Then, in order to improve localization accuracy, we propose a method of antenna selection. Finally, to deliver a real-time localization system we realize the proposed system by developing a software framework including a localization sever and a web-based location displayer. We have implemented the system on the commodity WiFi APs and the experimental results show that it can achieve accurate localization and has less the time cost.

Since that phase error exists in Channel State Information (CSI) measured by commodity WiFi device, the existing device-free tracking systems, based on Doppler extracted from CSI, suffer from the ambiguity of moving direction, which results in tracking error. Therefore, in this paper, we propose the design and evaluation of a device-free tracking system to solve this challenge. The contributions of this paper are listed as follows. Firstly, we creatively propose to use signal fading phenomenon, which is considered to be enemy to communication system, to estimate moving direction of the target. Secondly, we introduce ray tracing model to analyze the influence of moving target on received signal power and then build the profiles of received power of two cases, including the target moving towards and away from the link between transmitter and receiver. Thirdly, we rely on short-time Fourier transform (STFT) to process CSI, and then extract the power of Doppler in a time window. Then, Derivative Dynamic Time Warping (DDTW) is utilized to identify the moving direction. Consequently, the target can be tracked by moving speed and direction. We implement the proposed system on the commodity WiFi device and the experimental results show that it can achieve median tracking error 0.84m.

Multipath signals, which are suppressed in conventional localization algorithms, usually relate target location and structure of indoor environment through geometry parameters such as Angle of arrival (AOA) and Time of Flight (TOF). Thus, they can be exploited to locate the target and construct maps, which describe the structure topology of indoor environment. In this paper, we propose WalkAround, a multipath-assisted indoor localization and mapping system using the commodity WiFi signals, which contain phase errors caused by imperfect hardware and non-synchronized clocks. To realize accurate localization without interference of phase errors, we firstly construct a geometry model for jointly estimating the locations of target and scatterers which can be regarded as objects such as wall and furniture, by using TOF differences between the reflection paths and direct path. Then, with the help of AOAs we develop a locations searching algorithm based on Particle Swarm optimization (PSO). After that, we propose a density-based mapping algorithm with the locations of the scatterers and target, which does not need any anchor nodes or landmarks. We have implemented WalkAround in actual indoor environments by using the commodity WiFi devices. Based on the experiment results, the median location error of the target is 1. 49m and the constructed map matches the real structure topology of indoor environment well using only one receiver.

In indoor environment, multipath signals are rich and contain indoor geometry information, which can be used to locate targets. Based on this, a multipath-assisted indoor localization algorithm is proposed, which is different from the conventional localization algorithm regarding multipath signal as enemy. Firstly, differential Time of Flight (ToF) of multipath signals are used to construct the fitness function of Particle Swarm Optimization (PSO) with respect to the locations of target and scatterers. Then, PSO is used to jointly estimate the locations of target and scatterers, in which the AoAs of target and scatterers are used to determine location ranges. Secondly, to improve localization accuracy further, we propose a novel detection algorithm of bad scatterers based on the mutually exclusive characteristic manifested by the correct and wrong locations of scatterers. Then, Affine Propagation Clustering (APC) is used for all target locations estimated by scatterers to determine if the result of PSO is acceptable with the proposed criterion.

In this paper, we present a Real-time Interacting with Hand Description system via Millimeter-wave Sensor (Riddle) for human-computer interaction. Firstly, we describe a new approach to developing a radar-based system. When hand motions are captured by millimeter- wave radar sensor, the unique range information can be observed in the spectrogram. Compared to traditional hand gesture recognition systems based on optical sensors, the radar-based system avoids the influence of ambient light conditions. Secondly, we employ deep neural networks combined with connectionist temporal classification algorithm to recognize diverse hand gestures in real-time. Besides, we visualize the feature maps extracted from different layers to understand the deep neural networks. The deep neural networks are powerful to extract hand gesture features as well as class boundaries through a training process. Finally, we demonstrate that Riddle is capable of detecting six hand gestures and achieving high recognition accuracy of 96%.

The Internet of Thing (IoT) based applications can provide various services and be widely applied in intelligent home. With the tendency of house safety protection, the detecting accuracy and privacy of intrusion detection system, which detects the human motion in indoor environment, has become a continuing concern. Up to now, there are emerging many intrusion detection systems which employ different devices such as camera and infrared. However, the poor privacy and deployment of specialized devices are mainly disadvantages of the afore-mentioned systems for deploying in home environment. In this paper, we propose WLID, a whole-home level intrusion detection system based on RSSI (Received Signal Strength Indicator) measurements of WiFi in indoor complex environment. In order to expand the area of human presence detection, WLID cooperates with WiFi-enabled IoT devices such as smart TV, air conditioner and other smart devices. The detection system constructs a detection algorithm with the non-parametric statistical method by only using RSSI and realize whole-home level real-time detection by using software implementation. The experimental results show that WLID can achieve the consistent detection rate close to 100% in a practical home environment.

Knowing whether the propagation path between transmitter and receiver is Line-Of-Sight (LOS) or No-Line-Of-Sight (NLOS) propagation is a important factor for improving the performance of communication services and vast of mobile computing applications. Several promising systems in the current commodity WiFi networks analyze frequency- dependent amplitude and phase of Channel State Information (CSI) to achieve a LOS/NLOS recognition scheme, but robustness not be considered enough since there are many frequently-used wireless channels. With the development of Multiple-Input- Multiple-Output (MIMO), the spatial properties of channel can be obtained effortlessly. Consequently, it provides a potential to utilize the unchanged spatial information of multipath signals to get a robustness LOS recognition system. Here, we propose Wi-vision, an accurate and robust LOS recognition system based on Single-Input-Multiple-Output (SIMO) measurements in indoor environment. We creatively introduce the Hopkins statistic to measure the distribution of Angle-Of-Arrival (AOA) and relative Time-Of-Flight (TOF) of multipath under LOS and NLOS condition, respectively. The test results show that Wi-vision possesses consistent LOS and NLOS detection rate of above 91% and 87% respectively with different system configurations.