Published papers

This paper aims to address the issue of low through-put for users, caused by the random movement of High-Altitude Platform Stations (HAPS) due to winds. The proposed solution involves developing an Equal Clustering (EC) approach that groups users into high-density clusters, ensuring an equal num-ber of users in each cluster while maintaining low complexity. To further enhance the system's throughput performance, we fine-tune the antenna parameters using a Deep Q-Network (DQN) and the results of the EC clustering. To evaluate the effectiveness of the proposed method, we compare it with three Reinforcement Learning (RL)-based approaches and a K-Means clustering-based method. Simulation results indicate that both the EC method and the EC-aided DQN method successfully enhance the Cumulative Distribution Function (CDF) performance of throughput distribution when compared to the RL-based method for both rotation and shift scenarios. Furthermore, the EC-aided DQN method outperforms the K-Means clustering-based method in terms of the CDF of throughput performance.

A key concept in health monitoring systems for elderly people is the continuous and non-intrusive detection of their activities to identify when hazardous events such as sudden falling occur/are about to occur. The existence of obstacles in the environment largely limits the detection performance of existing approaches of activity detection relying on non-contact sensors. A simple, yet effective, approach to address this issue is the use of multiple sensors which collaborate with one another. In this paper, we propose an approach that relies on 2D Light Detection and Ranging (LiDAR) technology for activity detection. We employ multiple 2D LiDARs placed at different locations in a single room with difference obstacles (e.g., furniture) and working in coordination to construct a fuller representation of the activities being performed. Our approach transforms the concatenation of the different LiDAR data into a more comprehensible data format (i.e., images). The generated images are then processed using a Convolutional LSTM Neural Network to perform the classification. For 3 different tasks, namely activity detection, fall detection, and unsteady gate detection, our proposed approach reaches an accuracy equal to 96.10%, 99.13% and 93.13%, respectively.

With the aging of the population, the health of the elderly people has become a concern for many families. Monitoring of elderly people and the detection of their activities have become hot research topics. Several approaches have been proposed in the literature to detect such activity (e.g., falls), but their reliance on RGB cameras poses a serious threat to the privacy of the elderly. Therefore, Light Detection and Ranging (LiDAR) has attracted the attention of researchers. In this paper, we propose a novel method for fall detection using transfer learning and dynamic non-linear mapping to extract the human skeletons from depth images and implement a fall detection method using a Long Short-Term Memory (LSTM) neural network. We collected RGB images and depth images using 3D-LiDAR. The RGB images are used as a ground-truth for annotation and evaluation, whereas depth images are used as input to our proposed method. For fall detection, we achieved 98.5% accuracy, 97.0% precision, 100% recall, and 98.5% F1 score.

In recent years, non-contact biometric identification has garnered significant attention owing to its flexibility and its capacity to ensure privacy and confidentiality. Previous research has predominantly centered on harnessing cardiac radar signals detected by radar systems. This paper aims to compare two methods for person identification utilizing heartbeat signals extracted from radar data. In the first approach, we directly employ the heartbeat time series data as input for a deep learning model called InceptionTime dedicated to person identification. The second method explores the viability of using spectrograms, which are images generated from cardiac radar heartbeat signals, combined with other deep learning models, ResNet and a small convolutional neural network (CNN) model built by us, to achieve the same goal. Additionally, we assessed the robustness of both methods by introducing white Gaussian noise into the signals and evaluating their performance under different noise levels. For the second method, we implemented three image denoising methods: autoencoder, deep image prior (DIP), and Dilated-Residual U-Net (DRU-Net) to denoise the spectrograms after adding noise in the time domain. Experimental results reveal that the accuracy of person identification using spectrograms is 100% without adding noise based on the public dataset. The identification accuracy is the same as that using time series data. Following the introduction of noise, the accuracy of both methods decreases; however, the method based on spectrograms exhibits a more pronounced decline in accuracy compared to the approach using time series data directly especially for the small CNN model we built. The image-based method with DIP image denoiser shows outstanding performance when the value of noise is high compared with InceptionTime.

Heartbeat detection could enable various applications in the medical and health care fields. In particular, non-contact heartbeat detection can be acceptable for those who have difficulty wearing devices, such as burn patients and infants. A Doppler sensor could be a key device to realize heartbeat detection without any wearable devices. Many researches in recent years have focused on heartbeat detection using a Doppler sensor with a single beam Doppler sensor. However, when the SNR (Signal-to-Noise Ratio) of heartbeat components is low for the beam direction, the heartbeat detection accuracy is likely to degrade. In this paper, for more accurate heartbeat detection, we propose a heartbeat detection method based on beam diversity using a multibeam Doppler sensor. Through the preliminary experiments, we clarified that the SNRs of heartbeat components differ from one beam to another. This means that when the SNR is low for one beam, the SNR could be high for the other beams. Inspired by this fact, the proposed method extracts heartbeat components from all peaks detected by the multi-beam signals. To verify the heartbeat detection accuracy of our method, we conducted the experiments for different detection ranges. The obtained results show that compared to the conventional methods using a single beam, our proposed method using multibeam detected heartbeat more accurately. This indicates the benefit of exploiting beam diversity, i.e., the heartbeat detection accuracy can be improved by utilizing beam diversity.

Recently, non-contact blood pressure measurement is attracting increasing interests because it is suitable for the blood pressure measurement on a daily basis. A Doppler radar is a key device to enable the non-contact blood pressure measurement, and it can observe pulse waves due to aortic vasomotion. Specifically, PTT (Pulse Transit Time) is known to be correlated with the blood pressure, and thus the blood pressure can be measured by calculating the PTT from the pulse waves. However, to avoid the effects of slight body movements and breathing, the conventional method requires the subject to lie down on a bed and to stop breathing. In this paper, we propose a Doppler radar-based blood pressure measurement method that works against a subject with slight body movements and breathing. Due to respiration and slight body movements, the pulse wave could be distorted, which makes it difficult to calculate PTT. To solve this issue, in the proposed method, by relating features of the distorted pulse wave to those of the clean pulse wave using a deep learning model with LSTM (Long Short-Term Memory), the distorted pulse wave is transformed to a clean pulse wave that provides PTT correlated with the blood pressure strongly. The experimental results showed that PTT calculated using the proposed method has a strong correlation with actual blood pressure and the accuracy of blood pressure estimation is improved, compared with the conventional method.

Information exchange through a relay node is attracting attention for applying machine-to-machine communications. If the node demodulates the received signal in relay processing confidentially, the information leakage through the relay station is a problem. In wireless MIMO switching, the frequency spectrum usage efficiency can be improved owing to the completion of information exchange within a short time. This study proposes a novel wireless MIMO switching method for secure information exchange. An overloaded situation, in which the access nodes are one larger than the number of antennas in the relay node, makes the demodulation of the relay node difficult. The access schedule of nodes is required for maintaining the overload situation and the high information exchange efficiency. This study derives the equation model of the access schedule and constructs an access schedule with fewer time periods in the integer programming problem. From the computer simulation, we confirm that the secure capacity of the proposed MIMO switching is larger than that of the original one, and the constructed access schedule is as large as the ideal and minimum time period for information exchange completion.

The heterogeneous network (HetNet) is a promising network for beyond 5G to improve network capacity and spectral efficiency by developing low-power small cells within macro networks. Therefore, appropriate resource allocation schemes should be designed to balance between data rates and the cross-tier interference among different cells. To improve system robustness and also reduce the harmful interference to macrocell users (MUs), in this paper, we maximize the total interference efficiency (IE) (i.e., the ratio of the sum data rates of femtocell users (FUs) to the total cross-tier interference from femtocell base stations (FBSs) to MUs) with imperfect channel state information in a two-tier orthogonal frequency division multiple access based HetNet by jointly optimizing the transmit power of the FBS and the subcarrier allocation factor. Meanwhile, the outage rate constraint of each FU, the outage interference constraint of each MU, the maximum transmit power constraint of the FBS, and the subcarrier assignment constraint are considered simultaneously. The considered problem belongs to a mixed-integer non-linear programming problem and thus is NP-hard. The original problem with uncertain parameters is transformed into a deterministic and convex one solved by using the quadratic transformation approach, the variable relaxation approach, and Lagrange dual theory. The feasible region analysis and robust sensitivity are provided. Simulation results demonstrate that the proposed scheme has a lower interference to the MU and higher IE by comparing it with the traditional schemes.

The advantages of massive multiple-input multiple-output (MIMO) techniques depend heavily on the accuracy of channel state information (CSI). In frequency division duplexing (FDD) massive MIMO systems, the user equipment (UE) needs to feed downlink CSI back to the base station (BS) through the feedback link. The excessive feedback overheads and low reconstruction accuracy are the main obstacles for actual deployment of FDD massive MIMO systems. In recent years, deep learning (DL) has been widely used to address the above problems. In this letter, we propose a neural network by utilizing the self-attention learning and dense refine (SALDR), which improves the accuracy of CSI feedback. Furthermore, a unified decoder named SALDR-U is designed to realize different compression ratios for CSI feedback without changing any parameter. Simulation results show that the proposed SALDR and SALDR-U outperform the state-of-the-art network in terms of accuracy and overhead of CSI feedback. The source code for all the experiments is available at GitHub.The code of this letter can be downloaded from GitHub link: https://github.com/XS96/SALDR.

Change detection based on synthetic aperture radar (SAR) images is an important application in the remote-sensing technology field. However, the lack of labeled data has been a difficult problem in SAR image detection, especially for pixel-level change detection. In this letter, we propose a novel unsupervised change detection algorithm, which improves the detection accuracy by exploring features from both spatial and frequency domains of SAR images. In particular, first clustering is used as preclassification to obtain pseudo-labels and then by incorporating classifiers and pseudo-labels in terms of feature learning, a novel unsupervised detection algorithm is proposed. To improve the sensitivity of the algorithm to changed details and enhance the antinoise ability of the change detection network, the attention mechanism (AM) is integrated into the network to fully extract important spatial structure information. Moreover, a multidomain fusion module is proposed to integrate spatial and frequency domain features into complementary feature representations. This module contains multiregion features weighted by the channel-spatial AM and deep features filtered out by the gated linear units (GLUs) in the frequency domain. To verify the effectiveness of the proposed algorithm, it is compared against the other four SAR image change detection algorithms using three real datasets. The experimental results show that the proposed method outperforms the other four algorithms in terms of percent correct classification (PCC) and Kappa coefficient (KC).

Sarcasm is a sophisticated form of speech used to convey a message other than the apparent one. To date, there are numerous papers that have discussed the idea of automatic sarcasm detection and how it could be used for sentiment analysis improvement. The objective of this paper is to provide non-experts with a comprehensive overview of the state of research in this field and the main findings regarding sarcasm detection. Therefore, in this paper, we survey the state-of-the-art work done in this field, we recapitulate the research effort done, with focus on the more recent works, and we present the expected performance out of the proposed works. Nevertheless, we study in detail how this form of speech is used in different platforms, and how the way we express it evolves over time. We also discuss the proposition that suggests that sarcasm is a polarity switcher for sentiment analysis. To achieve these goals, we run some experiments on 3 different data sets, collected from 3 different platforms, and compare how sarcasm is employed in each. These platforms are Twitter, Reddit, and some news websites. Our experiments show that the way sarcasm is expressed is highly dependent on language mastery and the platform used. For instance, in the Twitter data set, whose users vary widely in age, language mastery, and understanding what sarcasm means, the overall precision of detection of sarcastic statement reaches 89.31%. In the reddit data set, the precision of detection of such statements is about 55.33%, and in the news data set, the precision reaches over 96.67%. Our experiments also show that, to a great extent, it is safe to affirm that sarcasm, when employed, switches the polarity of a given piece of text: for the 3 platforms presented above, sarcasm has been a polarity switcher for 89.3%, 89.1%, and 92.0% of their respective instances.

Federated learning (FL) has become an increasingly popular solution for intrusion detection to avoid data privacy leakage in Internet of Things (IoT) edge devices. Existing FL-based intrusion detection methods, however, suffer from three limitations: 1) model parameters transmitted in each round may be used to recover private data, which leads to security risks; 2) not independent and identically distributed (non-IID) private data seriously adversely affect the training of FL (especially distillation-based FL); and 3) high communication overhead caused by the large model size greatly hinders the actual deployment of the solution. To address these problems, this article develops an intrusion detection method based on a semisupervised FL scheme via knowledge distillation. First, our proposed method leverages unlabeled data via distillation method to enhance the classifier performance. Second, we build a model based on convolutional neural networks (CNNs) for extracting deep features of the traffic packets, and take this model as both the classifier network and discriminator network. Third, the discriminator is designed to improve the quality of each client's predicted labels, and to avoid the failure of distillation training caused by a large number of incorrect predictions under private non-IID data. Moreover, the combination of the hard-label strategy and voting mechanism further reduces communication overhead. The experiments on the real-world traffic data set with three non-IID scenarios show that our proposed method can achieve better detection performance as well as lower communication overhead than state-of-the-art methods.

In this paper, we focus on developing efficient multi-configuration selection mechanisms by exploiting the spatial degrees of freedom (DoF), and leveraging the simple design benefits of spatial modulation (SM). Notably, the SM technique, as well as its variants, faces the following critical challenges: (i) the performance degradation and difficulty in improving the system performance for higher-level QAM constellations, and (ii) the vast complexity cost in precoder designs particularly for the increasing system dimension and amplitude-phase modulation (APM) constellation dimension. Given this situation, we first investigate two independent modulation domains, i.e., the original signal- and spatial-constellations. By exploiting the analog shift weighting and the virtual spatial signature technologies, we introduce the signature spatial modulation (SSM) concept, which is capable of guaranteing superior trade-offs among spectral- and cost-efficiencies, and system bit error rate (BER) performance. Besides, we develop an analog beamforming for SSM by solving the introduced unconstrained Lagrange dual function minimization problem. Numerical results manifest the performance gain brought by our developed analog beamforming for SSM.

Objective: The main challenge in contactless heartbeat detection comes from breathing and/or body motion, which typically deteriorate heart rate (HR) measurements, due to incorrect selection of spectral peaks associated with HR. To acquire the reliable peak selection on spectrum within a relatively broad range, this article first proposes a spectral Viterbi algorithm. Second, a nonlinear source separation approach is further proposed to eliminate the noises generated by respiration and movements, suppressing the undesired spectral energy. Proposal: Inspired by the fact that the period of peak-to-peak intervals of heartbeat (RRIs) rarely vary within a short duration, a novel spectral Viterbi algorithm is proposed to estimate HR change, by the path metric (PM) of candidate paths of HR change. Moreover, based on a deep recurrent neural network (RNN), deep clustering (DC) is applied to separate out the targeted heartbeat source from Doppler signal, by dividing its spectrogram. Results: On the premise of wide-range HR measurement, the usage of spectral Viterbi algorithm substantially improved the precision compared with typical methods of HR estimation, both in the statuses of human subjects' sitting still and typewriting. In addition, the combination of DC obtains the smallest average errors. Significance: The proposed spectral Viterbi algorithm with DC is provided with three main strengths: 1) good adaptability to wide-range HR change; 2) robustness to nonlinearly mixed signal and noises; and 3) requirement of only a single-channel sensor.

Remote keyless entry systems have penetrated the vehicular market due to its convenience. However, its vulnerability against relay attacks exploiting the wireless channel is a serious threat. One of the prominent countermeasures is radio ranging between the key fob and the vehicle. Taking into account radio regulations, usability and power consumption among other things, a suitable approach is necessary to devise. This article proposes sub-GHz two-way ranging based on phase detection which offers high tolerance against reference clock offset between a pair of transceiver devices and a novel compensation scheme against multipath effects. We present the operation principle and theoretical performance under a two-path model of radio propagation and additive white Gaussian noise. In field experiments with prototype devices, the ranging accuracy is validated. We show that the proposed ranging produces sufficient performance for remote keyless entry systems, confirming if the key fob is within a few meters of the vehicle.

High altitude platform stations are attracting much attention as novel mobile communication platforms for ultra-wide coverage areas and disaster-resilient networks. Multi-cell configurations can increase communication capacity in a wide area. Conventional work optimizes the cell configuration for a multi-cell configuration based on a genetic algorithm (GA). This method identifies an optimal cell configuration in terms of spectral efficiency depending on the number of cells under a uniform user distribution. However, user distributions differ depending on location; thus, cell configuration optimization is also required for non-uniform user distributions. The problem is an increased number of parameters because each cell requires different antenna parameters for non-uniform user distribution compared to a uniform user distribution. This increases the difficulty of optimization, even when using a GA in some cases. Therefore, we propose a GA-based two-step cell optimization algorithm that comprises both search space reduction and antenna parameter optimization steps to address this problem. The proposed method employs the concept of co-evolution, i.e., a divide-and-conquer method. The proposed method divides multiple cells into several groups to reduce the number of optimized parameters and optimizes each group in order. In addition, the search space is reduced based on a marginal histogram obtained via optimization with a large step size. Simulation results demonstrate that this technique can reduce the number of combinations compared to the case without sub-area division. In addition, there are cases where the search space reduction further reduces the number of combinations without degrading the sum of the square root of throughput.

Objective: To monitor fetal health and growth, fetal heart rate is a critical indicator. The non-invasive fetal electrocardiogram is a widely employed measurement for fetal heart rate estimation, which is extracted from the electrodes placed on the surface of the maternal abdomen. The qualities of the fetal ECG recordings, however, are frequently affected by the noises from various interference sources. In general, the fetal heart rate estimates are unreliable when low-quality fetal ECG signals are used for fetal heart rate estimation, which makes accurate fetal heart rate estimation a challenging task. So, the signal quality assessment for the fetal ECG records is an essential step before fetal heart rate estimation. In other words, some low-quality fetal ECG signal segments are supposed to be detected and removed by utilizing signal quality assessment, so as to improve the accuracy of fetal heart rate estimation. A few supervised learning-based fetal ECG signal quality assessment approaches have been introduced and shown to accurately classify high- and low-quality fetal ECG signal segments, but large fetal ECG datasets with quality annotation are required in these methods. Yet, the labeled fetal ECG datasets are limited. Proposed methods: An unsupervised learning-based multi-level fetal ECG signal quality assessment approach is proposed in this paper for identifying three levels of fetal ECG signal quality. We extracted some features associated with signal quality, including entropy-based features, statistical features, and ECG signal quality indices. Additionally, an autoencoder-based feature is calculated, which is related to the reconstruction error of the spectrograms generated from fetal ECG signal segments. The high-, medium-, and low-quality fetal ECG signal segments are classified by inputting these features into a self-organizing map. Main results: The experimental results showed that our proposal achieved a weighted average F1-score of 90% in three-level fetal ECG signal quality classification. Moreover, with the acceptable removal of detected low-quality signal segments, the errors of fetal heart rate estimation were reduced to a certain extent.

Millimeter wave (mmWave) and massive multiple input multiple output (MIMO) are recognized as key technologies in the forthcoming beyond the fifth-generation (B5G) and the sixth-generation (6G) wireless networks. In this paper, a multibeam MIMO non-orthogonal multiple access (NOMA) scheme with weighted beam superposition for mmWave is proposed to enhance the system sum rate while ensuring fairness of users as much as possible. Specifically, a method of power allocation is adopted to guarantee the minimum demanded for the quality of service (QoS) of the weak users (lower channel gain)in each group. Furthermore, to improve further the sum rate after the QoS of the weak users is satisfied, the coefficient of strong users' beam gain are set to the largest value. In system simulations, we compare the performance of three multi-beam schemes, i.e, beam splitting, beam superposition and the proposed scheme, together with a single beam scheme, and a TDMA scheme at different levels of the SNR. The simulation results demonstrate that the system sum rate of the proposed method is much higher than TDMA scheme and competitive compared to the best scheme.