This research proposes a switchable non-physical LTE/WWAN antenna. It adopts the metal zone around the hinge slot, created by the existing display ground plane, keyboard ground plane, and metal hinges, to create the radiator. It then feeds a signal at the proper position in the hinge slot to effectively excite the resonance modes of the loop path around the hinge slot. Moreover, it adopts switchable feeding points to cover the LTE700/2300/2500 triple-bands required by LTE and the GSM850/900/1800/1900/UMTS penta-bands required by WWAN, respectively, and fulfils the operating requirements of fourth generation communication. In addition, since no additional deployment space and structure design for antenna is required, the design can resolve the space requirement for communication device, and provide new antenna design technology.

A wideband dual-circularly polarised (CP) substrate integrated waveguide (SIW) cavity-backed patch antenna is designed by using two-layer structure. The realised dual-CP, a cavity-backed square patch lies on the top substrate is excited by two pins, which are symmetrically located at an offset to the diagonal axis of the patch. These pins induce strong loading effect in the cavity and generate additional resonances in the vicinity of the patch resonant that help to achieve a wideband response by adjusting their locations. Thus, generation of dual-CP and wide impedance bandwidth is accomplished by using this new feeding topology. Moreover, the antenna is fabricated and tested for dual-CP. The experimental results show an impedance bandwidth of nearly 28.4% (8.59–11.43 GHz) and an axial-ratio bandwidth of 3.4% under the criterion <3-dB.

A novel meandered-line-loaded leaky-wave antenna (LWA) is proposed. Two half-mode substrate integrated waveguide-based LWAs are placed face to face for obtaining high peak gains, and the meandered lines are introduced to acquire larger ratio of the beam scanning range to the operating frequency bandwidth. It is demonstrated that the operating frequency band can be narrowed by almost 1 GHz due to the meandered-line structure under the condition that the scanning range keeps unchanged. The proposed LWA is fabricated and measured. Both simulated and measured results indicate that the designed antenna is a balanced LWA operating in frequency band 10.8–16 GHz (5.2 GHz centred at 13.4 GHz). Continuous beam scanning characteristic is observed from −60° to +34°, and comparisons are given out to validate the proposed design.

A novel configuration of a Fabry–Perot resonator (FPR) antenna for C band application, which produces a high gain, wider impedance and 3-dB axial ratio bandwidth is presented. Radiation element of the antenna is a square patch coupled with four points of crossed aperture. A feed network can change the rotation of the phase in aperture points and, by this technique, change the polarisation diversity (PD). This antenna is capable of variation circularly PD by changing the input port, also partially of the reflective surface which plays as the role of FPR resonator due to the structure and arrangement of cells capable of changing the PD. Experimental results clarify that the FPR antenna covers the impedance bandwidth of 4–6 for S11 ≤ −10 dB. Furthermore, the proposed antenna approximately has a constant gain from 4.4–7.5 GHz with a peak value of 18.1 dBic for each of the input ports. The antenna covers 3 dB axial-ratio bandwidth, between 4.35–7.85 GHz.

A polarisation-independent tunable absorber has been presented with wide tuning range. The proposed design consists of periodic patterns of square loops connected through varactor diodes, thus resulting in tunable performance. Due to four-fold symmetric design, the structure exhibits tunable characteristic for all polarisation angles unlike the earlier reported tunable absorbers. Furthermore, the configuration also eliminates the need for any external biasing network, and therefore avoids the design difficulties associated with the added bias grids. A prototype of the proposed structure has been fabricated, whose measured results show good agreement with the simulated responses under normal incidence.

A novel wireless dry EEG brain-controlled wheelchair with five steering behaviours is devised, which is based on two-stage control strategies combining sustained and brief motor imagery brain–computer interfaces (SB-MI-BCIs). Compared with the existing single modal brain–computer interface (S-BCI) controlled wheelchair, this new wheelchair has achieved more control commands and higher information transfer rate. Compared with their previous hybrid BCIs controlled wheelchair, subjects can control the wheelchair spontaneously and efficiently just by imagining the left hand, the right hand or both hands without any other visual attentions (e.g. P300 or SSVEP). The experimental results demonstrate that the SB-MI-BCIs can efficiently provide multiple independent control signals for wheelchair navigation.

The most widely used clinical technique for estimating bone movements in knee joints with acceptable accuracy utilises implantation of heavy metallic beads into the bones. These small metallic objects become visible as highly contrasting features in captured radiographs, thus facilitating post kinematic assessment of bones in human joints with pinpoint accuracy. Although, this approach provides necessary precision for clinical diagnosis, pre- and post-surgery planning for restoring joint function and locomotion, however, this approach also incurs high degree of invasiveness. In this study, a novel optical sensor is proposed for use in non-invasive motion measurement of knee joints in three-dimensional (3D) space. The approach is based on 3D angle-of-arrival (AoA) estimation using image processing procedures on the two-dimensional images of sensor's aperture shadows. The novelty of the sensor in terms of design, operation and its potential applications for a wide range of possibilities including precise AoA estimation for unconstrained 3D motion analysis of knee joints is described. Finally, error analysis of the sensor's AoA measurement uncertainty is provided for different incident angles of optical beams, which demonstrates the potential of the sensor not only for high-precision motion analysis of human joints; however, for other motion and position-aware systems.

Sleep apnea (SA) event occurs due to restraint in normal respiration. It requires accurate diagnosis, because of neurotic and cardiac disorders. In this work, particle swarm optimisation (PSO)-based Hermite decomposition algorithm is proposed, for identification of SA event using electroencephalogram (EEG) signals with parameterised classifier. The information from randomly varying complex EEG signals is extracted in terms of PSO optimised Hermite functions (HFs), with constraint of minimum error function. The Hermite coefficients computed from HFs-based statistical features are applied as input to PSO parameterised least square support vector machine classifier. The proposed decomposition for EEG signals provides negligible mean value of error function and obtain best results for identification of apnea event compared to existing methods.

Medical image fusion is the process of integrating two medical images with a visual enhanced single fused image, to attain a resultant image richer in information to aid medical practitioners in better diagnosis. A perceptual medical image fusion method is proposed by employing Internal Generative Mechanism. First, source images are divided into a predicted layer and a detail layer with a Bayesian prediction model. Then, the detail layer is merged with the energy of Tchebichef moments for blocks while the predicted layer is fused using the averaging strategy as activity level measurement. The fused image is finally obtained by merging coefficients in both fused layers. Experimental results prove that the proposed fusion algorithm is superior to the previously developed methods.

A 2.48 GHz class-C voltage controlled oscillator (VCO) without any dynamic back bias circuit is presented. The VCO uses three-path high Q-factor inductor as the low loss resonator and class-C cross-coupled nMOSFET pairs for high dc/RF conversion efficiency and direct cross-coupled pMOSFET for dc current reuse. At the supply voltage of 1.2 V, the core power consumption is 0.24 mW. The phase noise of the VCO is −124 dBc/Hz and the VCO figure of merit is −197.0 dBc/Hz. The VCO was implemented in the TSMC standard 0.18 μm SiGe BiCMOS process and occupies a die area of 0.799 × 0.809 mm².

The active vector effects on torque and stator flux linkage are different in direct torque control (DTC) system. These differences may cause the over-compensation or under-compensation to torque and flux linkage leading to large ripples. Aimed at making sure the system can work in an optimum state, a novel duty ratio modulation strategy minimising the torque and flux linkage ripples is designed. Every sector is divided into five small sectors, and a new switching table is designed. The impact factors and the active factors are first introduced into the DTC system. Subsequently, a novel duty ratio modulation method is derived. The superiority is verified through the simulation results.

Colour artefacts of demosaicked images are often found at contours due to interpolation across edges and cross-channel aliasing. To tackle this problem, a novel demosaicking method to reliably reconstruct colour channels of a Bayer image based on two different optimised mean-curvature (MC) models is proposed. The missing pixel values in green (G) channel are first estimated by minimising a variational MC model. The curvatures of restored G-image surface are approximated as a linear MC model which guides the initial reconstruction of red (R) and blue (B) channels. Then a refinement process is performed to interpolate accurate full-resolution R and B images. Experiments on benchmark images have testified to the superiority of the proposed method in terms of both the objective and subjective quality.

A bi-directional deformable block-based motion-estimation algorithm to improve the accuracy of a motion prediction in a 360-degree video, used for motion compensated frame-rate up conversion (MC-FRUC) is proposed. The block shapes are adaptively deformed to predict a motion based on the coordinate of the current block, considering the geometric property of the equi-rectangular projection of 360-degree videos. Experimental results demonstrate that the proposed technique can significantly improve the visual quality of the videos as compared with the conventional MC-FRUC.

Deep learning-based methods for stereo matching have shown superior performance over traditional ones. However, most of them ignore the inherent geometry prior of stereo matching when training, i.e. the reference image can be reconstructed from the second image in the visible regions. The reconstruction can be achieved by backward warping the second image using the disparity map of the reference image, while the visible regions can be calculated by left-right consistency check. This prior is useful especially when the ground truth disparity is sparse (e.g. the outdoor scene such as KITTI 2015). This prior incorporated into a two-stage end-to-end training process, both of which try to minimise the end-point-error with respect to the sparse ground truth disparity (supervised learning), and the reconstruction error (self-supervised learning). The predicted disparity and the reconstruction error of the first stage act as additional information, and are fed to the second stage to make further use of this prior knowledge to improve performance. Experiments on the challenging KITTI 2015 dataset show that the method improves the results in the foreground region, and ranks first among all the published methods on the D1-fg metric.

Natural image reconstruction based on compressive sensing (CS) has shown a promising performance in recent years. However, sometimes the restoration precision is not high enough. A novel CS algorithm using vector quantisation (VQ) error is proposed. First, the original image is compressed by VQ due to its extremely high compression ratio and strong ability to preserve details. Then compute the VQ error matrix and ignore the three least significant bits, which makes the error matrix much sparser. Next, to ensure a uniform distribution of sparsity of blocks, the error matrix is scrambled. Since the huge diversity among blocks has been largely reduced, they can be sensed with the same sensing matrix in space domain. At last, the reconstruction effect of the error matrix decides the total restoration performance. Experimental results have demonstrated the proposed method, at low measurement ratio, performs better in the aspects of perception and peak signal-to-noise ratio.

A multiple relays aided secret key generation algorithm is proposed to improve the generated secret key rate for Internet of things when the wireless channels change slowly. First, the two legitimate users and relays send training sequences in turn to estimate the channels. Then, the relays employ secure network coding technique to help the two legitimate users exploit the randomness of relay channels. Finally, the two legitimate users agree on a secret key without the help of relays. In addition, the optimal relays selection with optimal power allocation algorithm is present to further improve the secret key rate when relay selection is needed. The performances of the proposed algorithms are verified by Monte Carlo simulation results.

A Pd-functionalised hydrogen gas sensor based on an AlGaN/GaN heterostruture platform was investigated. The optimum bias voltage and reaction temperature were 0.1 V and 200°C, respectively, with which a wide range of hydrogen concentration from 0.1–4% in the air (i.e. 1000–40,000 ppm) was detectable. Not only low-power consumption but also high sensitivity with fast response was achieved due to the excellent properties of the AlGaN/GaN heterostructure platform. The sensitivity was 72.8% with a response time of ∼3 s in 4% hydrogen concentration.

An ultra-wideband asymmetric spiral-stacked Marchand balun fabricated in the standard 0.18 μm CMOS process is demonstrated. An innovative technique named in-phase current return path is incorporated into the design to enhance the balun's balance metrics. The proposed balun has a measured insertion loss <5 dB (3 dB for an ideal balun) and a return loss better than 10 dB from 20.8 to 60 GHz. Significantly, it provides a measured amplitude imbalance <0.3 dB and a phase imbalance <2° within 10–51 GHz range. The balun occupies an area of only 0.043 mm² without pads. The resultant highly balanced CMOS balun is shown as a high-performing millimetre-wave component suitable for the full Ka and/or Q band applications.

A novel decoupled empty substrate integrated waveguide (DESIW), which enables AC/DC decoupling in any device that is integrated in it is presented. The decoupling strategy is performed throughout a micro-milling square pattern of easy implementation that increases the insertion losses compared with the standard ESIW. However, the DESIW related losses are still tolerable, allowing the employment of the new periodic structure in many practical applications. In particular, it can be used for the design and manufacturing of reconfigurable devices which need a bias voltage on the whole device, or just on some of its particular areas. A broadband transition from DESIW to microstrip planar lines has been also successfully designed. The new line has been manufactured and measured.

A six-channel microstrip diplexer with compact size and high isolation is presented. The six-channel diplexer includes two pairs of multimode resonators with a pseudo-interdigital structure and a distributed coupling feeding line. Modal characteristics of the multimode resonator are explained by the even-odd-mode method. The centre frequencies of channel filters can be controlled independently, while bandwidths can be determined by the coupling coefficient and external quality factor. The distributed coupling feeding line has advantages of size reduction and small loading effect. Therefore, each channel filter can be designed independently. Moreover, the distributed coupling feeding line forms the zero degree feeding structure which can bring transmission zeros to realise high isolation. Finally, a six-channel diplexer prototype working at the centre frequencies of 1.8, 2.4, 3.0, 3.8, 4.7 and 5.8 GHz with over 30 dB isolation is fabricated and measured. Simulated and measured results show a good agreement.

Wideband high-power single-chip X-band class-F power amplifier with excellent output performances is reported. In order to achieve high-power class-F operation under minimised output capacitance impact at the frequency, high-voltage operation with minimised device size is pursued and its voltage swing becomes maximised by adopting high breakdown field-plate (FP) GaN high-electron-mobility transistor (HEMT) devices with cascode configuration. For the fabrication, 0.14 µm FP AlGaN/GaN HEMT technology with 58 GHz f_T and ∼100 V breakdown is used. Based on on-wafer test at 44 V Vdd, the fabricated class-F power amplifier delivers maximum 61.1% power added efficiency (PAE) with 9.2 W at 10.2 GHz. With the help of high-voltage operating cascode devices of minimised size, it also delivers excellent bandwidth of 1.6 GHz at the point of 50% PAE and high-power density of 4.42 W/mm is demonstrated.

A wide-band bandpass filter with high selectivity, compact size and wide stopband is presented in this Letter. This proposed filter uses dual-mode resonators (DMRs) as its basic tool. Stepped impedance technique is adopted in designing the DMRs for miniaturising its size. In the construction of filter design, three types of inter digital capacitors (IDCs) are employed; one is used for coupling between DMRs and other two are used at feed lines. The IDCs and open stub resonators are infused in the filter for curbing the harmonics and providing wide stopband. Input–output cross coupling is introduced by making use of hook-shaped feed lines for achieving good selectivity. A miniaturised size of 0.089λ_g × 0.10λ_g (29.2 × 33 mm²) with wide stopband up to 10.1f ₀ (f ₀ is the operating frequency of filter) with 21.5 dB harmonic suppression level is achieved with this design.

A novel three-dimensional high-isolated dual-band balun is proposed for the first time. A horizontal feeding plane, a vertically mounted plane, and four grounded resistors constitute the entire structure. By introducing the double-sided parallel-strip line with an inserted conductor plane to the middle of the vertical plane, the output matching and isolation, frequency-independent phase difference, and complete ground in the bottom layer of the horizontal plane are perfectly integrated in this novel balun. To prove this conception, a practical balun operating at 1.3/2.5 GHz is designed. As expected, remarkable consistency between the simulation and measurement is obtained.

Error-free transmission of 10 Gbit/s non-return to zero optical signals along 2.5 cm long silicon germanium waveguides at a wavelength of 1.98 μm is demonstrated. Bit error rate measurements confirm the absence of penalty during the transmission through a subwavelength of 1.3 and 2.2 μm wide waveguides.

Open-circuit-voltage (OCV) plays a significant role in state-of-charge (SOC) estimation for lithium-ion batteries. The slight difference in OCV at various temperatures can result in a large deviation of SOC estimation. In this Letter, a novel model based on Gaussian process regression is proposed to describe the sophisticated relationship among the OCV, SOC, and temperature. To validate the effectiveness of the proposed model, a comprehensive comparison with widely considered benchmarking OCV models is conducted. Experiment results demonstrate the proposed model can provide the most accurate prediction of OCV compared with benchmarking models.

A hybrid DC/DC converter is studied for direct current microgrid applications to have benefits of low switching losses with wide range of operation load, less winding turns of transformer secondary side, low ripple current on output side and high circuit efficiency. The studied converter includes a phase-shift pulse-width modulation full-bridge circuit and a half-bridge circuit to reduce the switching losses at lagging-leg switches. In the secondary side of the studied converter, two inductors, four diodes and two transformers are adopted to perform partial ripple current reduction and minimise the winding turns. The studied converters are verified with a 1200 W prototype by the experimental results.

Variable frequency controlled LCC resonant converter suffers from wide variation of switching frequency. Although fixed frequency controlled LCC resonant converter benefits from constant switching frequency, it suffers from narrow zero voltage switching (ZVS) range of power switch and high circulating energy. A pulse width modulation – pulse frequency modulation (PWM-PFM) hybrid controlled LCC resonant converter is proposed to achieve wide ZVS range of power switch with narrow variation of switching frequency. Experiment results of a 500 W prototype are provided to verify theoretical analysis of the proposed converter.

In this Letter, the author proposes a novel space–time adaptive processing (STAP) algorithm by enforcing sparse constraint on the beam-Doppler patterns for clutter mitigation when the number of training data is limited. By exploiting the sparsity of the beam-Doppler patterns of the STAP filter, the proposed algorithm formulates the filter design as a mixed l ₂-norm and l ₁-norm minimisation problem. Moreover, the proposed algorithm develops an adaptive approach to update the regularisation parameter. Simulation results illustrate that the proposed algorithm outperforms the traditional STAP algorithms in a limited sample support.

A practical deployment is presented for sonar imaging with fewer transmitters and receivers. Spatial coding via transmit beamforming is used for target detection and localisation. Transmit beamforming reduces the ambiguity in a model and improves imaging performance. To reconstruct the resultant underdetermined system of equations, a sparse signal reconstruction framework is used, and the inverse problem is solved by using an iterative optimisation algorithm. Deployment is optimised for better target detection and localisation performance. Successful sonar imaging reconstruction is guaranteed by the restricted isometry property with high probability. A numerical simulation shows that the proposed method can detect and localise targets efficiently, with better imaging results than those attained by conventional methods or the compressive sensing method without spatial coding.

Underwater acoustic multi-target tracking using bearing, Doppler and a conventional filtering method to determine the target location has problems such as low precision and inaccurate identification. Target echo time broadening is used first for the tracking process. The length of the target can be derived from the echo broadening, and the target length is obtained as a parameter by expectation–maximisation derivation, thereby increasing the dimension of the measurement for the estimation. Simulation results show that this method has higher tracking accuracy and target recognition accuracy than do other feature-aided tracking methods.

Lens-based millimetre-wave (mmWave) massive multiple-input multiple-output (MIMO) can utilise beam selection to reduce the number of radio-frequency (RF) chains. However, most of the existing beam selection schemes involve high complexity, especially with the fast time-variant mmWave channels. In this Letter, by exploiting the mmWave channel property that the angles of departure of channel paths are slowly varying, the authors propose an adaptive neighbourhood search (ANS) beam selection. The key idea is to utilise the concept of neighbourhood search developed from machine learning to select the beams with significantly reduced complexity, where the neighbourhood range is adaptively adjusted to avoid the local optimum. Simulation results show that the authors’ scheme can achieve the performance close to the conventional real-time beam selection schemes.

To design a dynamic correlated channel simulator which can precisely depict the space–time correlation properties and be suitable for the massive multi-user multiple-input multiple-output (MU-MIMO) system, a 2D autoregressive (2D AR) model-based method is proposed. Specifically, by exploiting stationarity in space–time domain, the authors use 2D AR method to create the space–time channel matrices of every user separately. In this way, the space–time dimension of the simulated channel can be modified flexibly without rebuilding the 2D AR model, and spatial correlation matrix with excessively big size is not necessary to guarantee the accuracy of the proposed simulator, making it suitable for massive MU-MIMO system. Simulation results verify the advantages of the proposed simulator.