Search

Sign language is the primary communication medium for the deaf-mute community. However, the literacy of understanding and using sign language is hard to gain without professional training. In this paper, we explore the use of low power frequency modulated continuous wave radar for automatic sign language recognition. The proposed system is composed of a radar, a sound cluster and a computer for transforming signals to spectrograms. Furthermore, as the time-frequency spectrograms are high-dimensional data with redundant information, we then perform dimensionality reduction by extracting the histogram of oriented gradients features from these spectrograms. The features are finally classified by the k-Nearest Neighbour algorithm and a classification result of 95.8% is achieved on the five testing signs. The impact of the k value in the k-Nearest Neighbour is also investigated.

For polarimetric synthetic aperture radar (PolSAR) images, building extraction has been a challenging topic for long time in applications of land-use and land-cover analysis. Due to similar structures of buildings and such vegetation as forest, they often exhibit similar PolSAR scattering characteristics that are often difficult to distinguishing. Recently, deep Convolutional Neural Network (CNN) has been widely investigated for image processing with many promising results. This paper proposes a method that combines polarimetric features with the CNN network to realize the comprehensive utilization of polarimetric and contextual information of PolSAR data for the extraction of building areas in PolSAR images. Comparison experiments on both ESAR and EMISAR L-band PolSAR datasets show that the proposed method can generate better results for building extraction.

Object-Based Image Analysis (OBIA) is one of the cutting edge image processing technologies attracting significant attention in remote sensing industry recent years and Synthetic Aperture Radar (SAR) possess a high potential for classification of earth surface. In this paper, a novel classification method, integrating OBIA and majority voting strategy, is proposed for land-cover classification based on SAR data. The main purposes of OBIA are delineating objects, extracting feature and measuring similarity by using superpixel segmentation technology and cosine similarity algorithm. The color histogram is utilized as feature in this study. Majority voting strategy provides an effective way for class determination. At last, the experimental result is presented and it demonstrates the proposed classification scheme is an appropriate method for land cover classification of SAR data.

This paper presents a comparative study of different methods for cancellation of surface reflections in ground penetrating radar (GPR) technology. Mean subtraction and singular value decomposition are studied and compared using the same experimental data set from real soil medium. Besides, a new method named dual-band cancellation method is also presented in the comparison. Experimental results show that the conventional surface cancellation methods have their own drawbacks when the medium surface is not flat. The newly proposed method has advantages in soil surface clutter cancellation.

There are great interests to study detection, and classification of pedestrian and cyclists using low-cost frequency modulated continuous wave (FMCW) radar sensors for intelligent transport systems (ITS). This paper presents a study about detection and classification of pedestrian and/or cyclists based on realistically simulated FMCW radar sensor echoes. From radar echoes, the micro-Doppler signatures of different targets can be observed through time-velocity diagrams and then extracted. Using a relatively small number of data samples for training, an Extreme Learning Machine (ELM) can be trained to detect the number of targets and classify their types. Numerical experimental results show the effectiveness of the proposed approach and the potential of low-cost radar sensors in ITS applications.

Crop classification is one of the focused topics in remote sensing study nowadays. Optical imageries, although providing many information, are often contaminated with cloud or other weather effects while SAR imageries are more resilience to those. Temporal series data are often used to improve classification accuracy, especially in crop classification. This paper investigates the usage of temporal series SAR imagery on crop classification in vast rural area of China. The selected area of interest has a complicated, heavily mixed agriculture as well as lots of non-agriculture landcovers. Total 9 classes are considered, 6 of them being crop types. A pixel-based classifier using subspace k-th Nearest Neighbor (KNN) algorithm is applied to open source Sentinel-1 polarized SAR data. The study results in an overall accuracy of 97.23% for all 9 classes in 5-fold cross validation, indicating a promising application for agricultural monitoring using SAR data.

This paper presents a progressive morphological filtering (PMF) approach on land cover clustering based on Synthetic Aperture Radar (SAR) remote sensing image. For SAR image containing only intensity and texture information, pixel-based classification methods produce less satisfactory results due to speckle noise in SAR image. While morphological filtering operation can be used to improve such result. In the classification step of the SAR image, K-means clustering method is used in combination with Decision-Tree method for classification of the clusters. The proposed PMF approach can be used within the classification steps to not only reduce classification errors caused by speckle noise, but also improve the clustering visualization effects with flexibility in meeting various purposes, such as land cover overview for land surveying or planning. Two SAR images of Singapore are used to validate the performance of the approach.

This paper presents a real life experimental study on moving human target localization using dual frequency continuous wave (DFCW) radar combined with Doppler Direction-of-Arrival (DDOA) algorithm. The DDOA method was adopted for moving target localization in both open-space and through-the-wall scenarios, which estimates target location based on range information and DOA information derived from Doppler phase. DOA estimation method taking into consideration of phase delay instability of RF cables is presented to eliminate cable effect. Experimental results for both scenarios are conducted and results show that good performance can be obtained for open-space trials but limited performance is observed for through-the-wall trials due to limited SNR. However, target can still be tracked to some degree in the latter case.

This paper presents a novel algorithm for designing beam-space transformation with robust interference suppression, reducing the computation complexity, and improving the performance of applied signal processing algorithms. This approach constructs an objective function that addresses multiple performance criteria simultaneously including the in-sector estimation accuracy and the out-of-sector interference rejection capability, and then utilizes a numerical optimization paradigm to search for the solution. The algorithm is flexible, robust and data independent, which performs well in harsh environment involving a small number of snapshots and highly correlated/coherent sources. Simulation results confirm its effectiveness in the presence of strong moving interference.

This paper presents a simple 4-port directional antenna as a switchable antenna or MIMO antenna with enhanced gain and interference suppression for V2X communication. A special reflecting wall in cylindrical shape is introduced to the center of the antenna. This wall enhances the gain of the antenna element in one direction while also provides space and flexibility for other antennas. This structure provides a new degree of freedom for optimizing the antenna performance. Simulation and measurement results show that the antenna can perform well in DSRC band.

The recently developed event cameras can directly sense the motion in the scene by generating an asynchronous sequence of events, i.e., event streams, where each individual event (x, y, t) corresponds to the space-time location when a pixel sensor captures an intensity change. Compared with RGB cameras, event cameras are frameless but can capture much faster motion, therefore have great potential for recognizing gestures of fast motions. To deal with the unique output of event cameras, previous methods often treat event streams as time sequences, thus do not fully explore the space-time sparsity of the event stream data. In this work, we treat the event stream as a set of 3D points in space-time, i.e., space-time event clouds. To analyze event clouds and recognize gestures, we propose to leverage PointNet, a neural network architecture originally designed for matching and recognizing 3D point clouds. We further adapt PointNet to cater to event clouds for real-time gesture recognition. On the benchmark dataset of event camera based gesture recognition, i.e., IBM DVS128 Gesture dataset, our proposed method achieves a high accuracy of 97.08% and performs the best among existing methods.

Smart Antenna System (SAS) is one of the latest hot topics in both academic and industry. It is regarded as the antenna solution for the future generations of wireless communication system. Phased array is widely used in SAS. However, problems at size, cost and meeting the industrial regulations, for example, the sidelobe level, remains. Current studies in array pattern optimization is limited to linear array with one dimension in polar space. Such optimizer is not sufficient to cope with current complex array designs. In this paper, a design of 3D pattern optimizer using popular optimization algorithm, namely Particle Swarm Optimization (PSO), is presented. The renovated optimizer is put to tests on a complex array setup, which is a dual-layer circular array with center element.

A low-profile top-hat shorted monocone antenna with ceramic loading is proposed. SiO 2 , a kind of ceramic, is first employed as matching layer to improve antenna's matching property and support the antenna's structure. Besides, for further increasing the bandwidth of monocone antenna, top hat and shorting pins are applied to reduce the lowest operating frequency and thus decrease the thickness of the monocone antenna to O.068λ Low . Simulated results show that a 6:1 bandwidth can be obtained.

Automotive radar sensors play critical roles in self-driving that would revolutionize transport and shape the society. How to process automotive data and sense the environment has become a study of concern. This paper presents a study of objective classification and tracking based on a set of measured data from 4 high performance 77 GHz radar sensors mounted on a vehicle moving in a complex environment. The radar data is processed dynamically in frames with integrated three processing techniques i) the clustering processing based on DBSCAN clustering method, ii) the tracking processing based on Kalman Filter, and iii) the classification processing by studying the features of clusters of different categories of road users.

Intersatellite Data Relay System (IDRS) is a critical technology for emerging satellite communications. It allows a constellation of satellites to form a network on themselves and hence gains constant link to & from the ground. CubeSat is the cutting-edge concept in commercial satellite industry. It features a miniatured size with multiple units of 10×10×10 cm3blocks. To realize IDRS on a CubeSat is a major challenge in both industry and academic. Compact electrically steered smart antenna is the only solution. This paper presents a design of a circularly polarized (CP) antenna array together with its RF frontend flowchart. The smart antenna system achieves small size by utilizing end-firing CP antenna as its element. Simulations of different application scenarios are shown. Concerned specifications such as G\T and sidelobe level are taken into account. The design concept is found suitable for both space and personal mobile satellite ground terminals.

This paper presents a novel integrated multiband vertically polarized antenna based on two half Vivaldi antennas and a slot antenna for V2X communication. It covers multiple frequency bands for vehicular communications, including LTE, WiFi-2.4G, WiFi-5G, Bluetooth, and DSRC bands. The integrated design can reduce mutual coupling among 3 ports and 5 application bands and can be fabricated a single piece FR-4 laminate using low cost PCB technology. The antenna prototype has been fabricated and tested with satisfactory results.

A novel low profile wideband monocone antenna with low cost printed circuit board (PCB) fabrication technology is proposed. Step structures are proposed to simulate the cone structure of traditional monocone antennas, based on three stacked layers of PCB with equal height. The monopole-like radiation patterns and the input reflection coefficients below −10dB are achieved from 2.06GHz to 7.36GHz. Techniques like applying the shorting pins and the annular ring are used to decrease the lowest working frequency. The thickness of the proposed antenna is reduced to only about 0.06λ 0 at the lowest frequency. The ripples in the H-plane pattern is always below 0.5dB.

Active direction of arrival estimation system can significantly improve the degrees of freedom (DOFs), since the vectorized covariance matrix of the data measurements from transmit and receive arrays can be considered as observations from a augmented virtual array, whose elements are given by the difference co-array of the sum co-array (DCSC). This article presents a generalized methodology to design the high performance active non-uniform linear arrays (NLAs) after performing detailed analyses of bridges between active NLAs and passive NLAs. Closed-form solutions in terms of positions of physical/virtual sensors and DOF capacities are derived. Simulation results validate the effectiveness of the proposed method.

Adaptive digital beamforming is widely used in modern radar and wireless communication systems. Evolutionary algorithms (EA) are powerful tools for more flexible adaptive beamforming. The major drawback of conventional evolutionary algorithms is generally slow in searching for solutions. Bat Algorithm (BA) is a relatively new and faster evolutionary algorithm. In this paper we apply Bat algorithm in adaptive beamforming and also compared its computational efficiency with that of Particle Swarm Optimization (PSO) algorithm which is considered much faster than Genetic Algorithms (GA). Numerical experiments show BA is very efficient, showing a promising tool for adaptive beamforming.

This paper reports recent works on calibrating peak power sensors with the measurement traceability to a primary power standard, microcalorimeter. Several practical measurement setups have been evaluated and compared. Those uncertainty components relevant and significant to the calibration process are also discussed.

This paper reports recent theoretical modeling and analysis of traceable peak envelope power measurements for mobile communications. Double Sideband Large Carrier (DSB-LC) amplitude modulation has been used, for establishing the measurement traceability to primary RF and microwave power standard. In this work, the reference peak envelope power calculated from modulation index and reference average power is used in calibrations. This is followed by performance comparison of the proposed model with detailed uncertainty analysis.

This paper presents a study of chipless RFID design and detection with adaptive direct path cancellation. Chipless RFID is an attracting topic but its reliable detection has been a problem in applications. In this study, we introduce and employ adaptive direct path cancellation to process the raw data received with significant improvement of frequency signature detection and reliability compared to the results by others.

A variation of bowtie printed dipole antenna is proposed and designed for a digital beamforming (DBF) array in Digital Video Broadcasting-Terrestrial (DVB-T) based passive radar applications. The proposed sleeve bowtie dipole reflector antenna element has a wider beamwidth/bandwidth than that of conventional bowtie/half-wavelength dipole reflector antenna, providing useful wider field of view and wider bandwidth for the DVB-T based passive radar applications. The proposed antenna has a bandwidth of 20.6% over the DVB-T band and an E-plane beamwidth 73° with more than 20% beam broadening effect.

The substrate integrated waveguide (SIW) allows the construct of some types of planar antennas which cannot be conventionally integrated on the substrate. However, owing to some technology restrictions, designed SIW horn antennas normally work in the frequency above 10 GHz. This paper proposes a 6.8GHz low-profile H-plane horn antenna based on the ridged SIW, which allows the substrates thinner than λ/10. Far field radiation patterns are reported to reveal the good performance of proposed antenna. Comparisons among the ridged SIW horn antennas with different number of ridges are also made to shown the matching improvements.

This paper presents a study of Synthetic Aperture Radar (SAR) imaging based on random sparse array (RSA) and compressed sensing (CS) for target detection. Extensive numerical experiments based on real RSA measurement data are carried out and statistical analysis is performed to compare the performances of the CS technique and back-projection (BP) method using RSA. Performance analysis of the results show that RSA-CS method has better performance than the RSA-BP method and can achieve acceptable performance with much smaller data amount than full SAR imaging.

This paper presents a study of various types of radar array architectures for low cost radar antennas with digital beamforming (BDF) capabilities. In addition to popular oneway and twoway beamforming, the emerging synthetic array radar (SAR)-like DBF and multiple-input-multiple-output (MEMO) DBF are examined with beamforming equations. New simple approximation formulae about half-power beamwidth (HPBW) estimation for SAR-like array and MIMO array are proposed for easier design reference. The comparison study shows that the SAR-like DBF or MEMO DBF can achieve much better angular higher resolution than the conventional oneway or twoway beamforming techniques.

Adaptive beamforming is critical in high performance radar and wireless communications applications. This paper presents a study of adaptive beamforming for arbitrary array by applying the powerful and versatile particle swarm optimization (PSO). Numerical experiments for sidelobe suppression, nulling, null steering and array failure correction have demonstrated that the presented PSO approach is effective for adaptive beamforming.

Ground Penetrating Radar (GPR) with Ultra-Wideband (UWB) waveform has wide applications in numerous fields. Most of them utilize the conventional sensors configuration, which is either Single-Input Single-Output (SISO) or Single-Input Multiple-Output (SIMO). In this paper, we explore several potential applications of UWB GPR with Multiple-Input Multiple-Output (MIMO) configuration. Three sets of simulation schemes are investigated. The simulation results show several improvements provided by MIMO array configurations over conventional array configurations on the corresponding GPR applications.

This paper presents a MIMO cross array with azimuth-elevation combined beamforming. Compared to an M×N rectangular array with one-way and two-way beamforming, the number of required array elements could be reduced from M×N to M+N+4 so as to drastically reduce the system cost while maintain higher angular resolution performance. Compared to an M+N cross array with one-way and two-way azimuth-elevation combined beamforming, the proposed MIMO cross array could achieve higher angular resolution with 4 additional transmitting antennas. This MIMO cross array with azimuth-elevation combined beamforming is designed for DBF systems for detecting single target or multi-target in different ranges (only when targets could be separated from each other after range detection processed) and it can greatly reduce the usage of array elements as well as other RF components.

This paper presents a novel multi-resolution composite array (MRCA) concept for high resolution digital beamforming (DBF) with much less elements than conventional uniform linear array. MRCA examples are presented and compared to conventional full-filled uniform linear arrays and multiplicative linear arrays to show the novelty and effectiveness.

This paper presents a novel design of a wearable textile antenna for 2.4 GHz ISM bands wireless applications. This design is an L-shape patch antenna with coplanar waveguide (CPW) feeding structure. The presented design is able to achieve wider bandwidth and larger tolerance necessary for wearable applications. The validation of the design is shown by comparison of realistic field simulation with measurement results of an antenna prototype. It is shown that the return loss response has negligible variations when placed on body which makes the presented design a promising candidate for wearable applications.

Vector sensors are made of electric dipoles and magnetic loops. A full six vector sensor can measure the complete six electric and magnetic field components at one point in the space. Vector sensor concept was proposed from signal processing community and associated direction of arrival (DOA) estimation algorithm was based on isotropic radiation model. This paper shows that the radiation pattern compensation of sensor element is necessary for practical DOA application with improved accuracy, from more than 12° to about 1° for a realistic signal of 10 dB SNR. As a useful example, we consider a simpler yet more popular low-grazing angle application with a reduced vector sensor of two magnetic loop sensor elements for DOA estimation of a low-grazing angle source with vertical polarization. The reduced vector sensor can greatly simplify the system design, fabrication, measurement, and data processing.

In this paper, the emerging multiple-input multiple-output (MIMO)-array configuration, where the transmitters and the receivers are correlated, is introduced for ultra-wideband (UWB) ground penetrating radar (GPR) imaging and compared with conventional single-input single-output (SISO)-synthetic array radar (SAR) imaging configuration, where each pair of sensors works independently. Several migration algorithms are studied and compared for both UWB SISO-SAR GPR and UWB MIMO-Array GPR configurations. The comparison study is based on six sets of realistic simulation results and analysis.

Many research results on biological effects of cellular phone radiations agree that the main possible hazards are the thermal effects while some other results reveal that the thermal effects due to handsets radiation are negligible. Few can deny the fact and puzzle that some users did suffer from such symptoms as headaches due to cellular phones. This paper provides a potential cause or possible answer to the puzzle through analysis and understanding of GSM standards. We can also reveal that the low power low duty cycle pulsed train amplitude envelop modulation of GSM uplink signals is the main cause of such non-thermal biological effect for non-ionizing radiation.

Vector circular array (VCA) is a relatively new array configuration for signal direction and polarization estimations. The objective of this study is to find a more accurate and robust DOA and polarization estimation approach for VCA. Fourth-order cumulants based MUSIC (FOC-MUSIC) method is attractive for possible high resolution in both white and colored Gaussian noise. This paper presents a study by comparing the FOC-MUSIC with conventional MUSIC for both white and colored Gaussian noise. Through stochastic simulation and error analysis, it is found that the adopted FOC-MUSIC can achieve more accurate estimations of direction and polarization in both white and colored Gaussian noise, especially for low SNR scenarios.

A microstrip grid array antenna is designed at 24 GHz on a single-layer RT/duriod 5880 substrate with a thickness of 0.31 mm. Simulated results show that the microstrip grid array antenna achieved a 10-dB impedance bandwidth of 210MHz, a 3-dB gain bandwidth of 500-MHz, and the maximum gain of 30.1 dBi at 24.15 GHz.

Synthetic Aperture Radar (SAR) is a popular and important high resolution radar imaging method. Compressed sensing is an emerging technology with potential to reduce high data acquisition, storage, and transmission of conventional SAR. In this paper, we present a study about SAR imaging with compressed sensing and random aperiodic array. The experiment results show that the proposed approach is effective in SAR target detection.

An innovative low-cost and higher than conventional angular resolution digital beamforming (DBF) array radar system is proposed and introduced in this paper. The proposed DBF radar system uses only one set of transceiver plus an additional microwave switch. Compared with conventional configurations with N sets of expensive microwave digital transceivers, the proposed concept can drastically reduce the system cost. Besides, we also propose an innovative DBF scheme that has narrower beamwidth the the conventional one-way or two-way beamforming configurations. The total system cost reduction factor is up to 2 times the number of array elements.

A microstrip grid array antenna that radiates a pencil beam is reported. The prototype was designed and fabricated on RT/duriod 5880 substrate with the physical size of 40×40×0.787 mm3 and the number of 16 radiating elements. Measured results show that the microstrip grid array antenna achieved 4.1% impedance bandwidth and 15.5-dBi gain at 24.25 GHz.

This paper presents a study of a battlefield simulation platform with consideration of sensor network and optimal resources management. This development is mainly based on two public ally available software - GECCO and REPAST. The platform allows integration with a user developed optimization engine for optimal battlefield management. A simulation example was presented to show the application to a wireless sensor network composed of unattended ground sensors and unmanned aerial vehicles for target detection, classification and localization in the battlefield.

This paper presents a novel low-cost time-division multiplexing MIMO radar system concept by using only one transmitter and one receiver. For an MIMO radar system of M-element transmitting array and N-element receiving array, the number of required transmitters/receivers is reduced from N+M to 2 for considerable cost reduction.