A quad-band filtering antenna based on a slotted half-mode substrate-integrated waveguide cavity is proposed. Three cavity edges are shorted by metallic posts and one edge is opened. Then three rectangular slots are etched to obtain multi-band operations and coaxial feeding is employed. Four radiation frequencies with similar field distributions but different resonating size can effectively radiate cavity energy through the opened edge and etched slots, resulting in good in-band antenna features. On the other hand, three radiation nulls whose dominant field is concentrated around the feeding port can generate little energy leakage, leading to good out-of-band filtering characteristics. The proposed design has been fabricated with an overall size of 0.40λ₀ × 0.20λ₀. Measured radiation frequencies are at 3.35, 3.88, 4.54, and 5.18 GHz with high gain of about 5.0 dBi and efficiency of 70.0%, whereas measured radiation nulls are at 3.54, 4.24, and 4.91 GHz with very low gain of −16.75, −13.46, and −12.06 dBi and efficiencies of 0.59, 0.88, and 1.42%. Moreover, unidirectional patterns with stable radiation are experimentally confirmed.

Accurate prediction of intended gait speed is crucial for facilitating volitional muscle activity during robot-assisted gait training. Considering that most patients find it difficult to perform gait at relatively high speeds, it is also imperative that the prediction model performs well at slow gait speed. Furthermore, ample time for signal processing should be given prior to robot assistance such that the patient is not subjected to unwanted resistance or lag. Here, the authors propose to predict the intended speed of the contralateral limb via kinematics of the current swing limb. Three lower-limb positions were evaluated using an inertial measurement unit sensor during treadmill gait sessions. The results suggest that the authors’ model performs well even during slow gait speeds with values as high as 0.86.

The sigmoid activation function plays an essential role in implementing neural networks in hardware. However, due to a high degree of non-linearity, the hardware implementation of the activation function itself is a daunting job. In this Letter, the authors proposed an accuracy controlled iterative algorithm for the piecewise linear approximation of the sigmoid function. The proposed algorithm approximates sigmoid activation with an optimised number of linear functions for a given error. The accuracy and hardware complexity of the resulting approximation function are compared with the state-of-the-art literature.

A background calibration technique is proposed to correct bit weights in pipelined-successive-approximation-register (SAR) analogue-to-digital converters (ADCs). By splitting the second stage, the input signal interference is mostly removed, thereby greatly enhancing the convergence speed of the algorithm. Besides, the dither signal assists to eliminate mismatch issues between the partially split ADCs, thus relaxing the analogue overheads. According to the simulation, after calibration, the spurious-free-dynamic-range and signal-to-noise-and-distortion-ratio are improved from 53.2 to 88.2 dB and 49.5 to 75.2 dB, respectively. The calibration algorithm converges with about only 600 K samples.

This Letter presents an improvement of disparity map refinement stage using adaptive least square plane fitting technique. This technique is proposed to increase the accuracy on the final stage of stereo matching algorithm. Fundamentally, the accuracy of matching process depends on the robustness of an algorithm on the plain colour region, depth discontinuity and repetitive pattern area. These regions are difficult to be matched and very challenging. Thus, this Letter proposes multiple point selections of disparities on a plane to adaptively group similar disparity values. Then, the distance of each disparity is calculated to ensure the values are accurately grouped. The distance refers to an assigned reference radius disparity location. Based on the experimental results using the KITTI and Middlebury benchmark images, the proposed method is capable of increasing the accuracy on the challenging regions and the quantitative measurement.

A novel deep-learning-based method for semantic segmentation of RGB and thermal images is introduced. The proposed method employs a novel neural network design for multi-modal fusion based on multi-resolution patch processing. A novel decoder module is introduced to fuse the RGB and thermal features extracted by separate encoder streams. Experimental results on synthetic and real-world data demonstrate the efficiency of the proposed method compared with state-of-the-art methods.

Digital multitoning is an advanced form of halftoning-based printing technique which renders multiple tones at each pixel of an image. This superior scheme delivers an improved print quality in terms of smooth texture, homogeneity and the suppression of artefacts. This Letter examines the potential scope of multitoning as an effective strategy to embed a binary watermark within a multitoned image. The proposed method uses the significant difference among the highpass wavelet coefficients achieved by discrete wavelet transform for watermark embedding. Furthermore, an improvised ordered dithering technique using adaptive threshold quantisation is also incorporated, which not only makes the proposed method time-efficient but also offers uniformity to the final image. The proposed watermarking technique is imperceptible and yields watermarked images with high peak signal-to-noise ratio value. Experimental simulations demonstrate that the proposed watermarking method is superior to the state-of-the-art methods in the context of both objective and subjective assessments.

In this Letter, a multiscale kurtosis maps fusion method is proposed for small infrared target detection. The authors have made two main contributions: first, they propose the scheme of using multiscale kurtosis maps to match the shape and the size of the small target; second, they present a dot product operation to achieve the fusion of multiscale kurtosis maps. Experimental results show that the presented method has a higher background suppression factor and signal-to-clutter ratio gain values than that of compared baseline methods, and it works effectively and more stably in complex backgrounds.

While promising progress has been achieved in the zero-shot learning (ZSL) task. The existing approaches still suffer from the strong bias problem between the unseen and seen classes. This Letter presents a unified feature generating framework equipped with a boundary decision loss to tackle this issue in ZSL. Specifically, the hybrid semantic and visual classification strategy is proposed, which can effectively align the bidirectional visual-semantic interactions. Furthermore, this Letter introduces a decision loss that optimises the decision boundary of seen and unseen classes to further alleviate the confusion of generated features. Extensive experiments on three popular datasets animals with attributes, Caltech-UCSD-Birds 200-2011, and SUN show that the proposed approach outperforms previous state-of-the-art works under both traditional ZSL and challenging generalised ZSL settings.

A design of 4/6 intertrack-interference (ITI)-mitigating error correction modulation code for bit-patterned media recording systems is proposed in this Letter. Generally, coding schemes in data storage devices are usually accomplished using a combination of error-correction coding and modulation coding. However, an interesting approach is that the modulation code itself has the capability of correcting errors. Such a modulation code design will reduce the total size and complexity of the data storage devices. In this Letter, the authors design a new 4/6 modulation code, where the code not only mitigates the effect of ITI but also maintains minimum Hamming distance of 2 to build up a trellis structure providing error correction capability of the proposed code. Simulation results show that the proposed design outperforms prior arts. For instance, the performance of the proposed design is better than that of a conventional 4/6 modulation code about 8 dB at a bit error rate of 10⁻⁷ under 20% read-head offtrack for the area density of 4 Tb/in².

This Letter presents a dual-mode substrate integrated coaxial line (SICL) based bandpass filter for -band application. The proposed bandpass filter is realised by exciting the even and odd modes in an SICL ring resonator. A novel perturbation technique to control the coupling between orthogonal modes is realised by varying the width of the outer conductor formed by metallic vias in this planar coaxial line. Further, the effect of perturbation length and inset on resonant frequency and attenuation pole is studied. The two transmission zeros produced close to the pass band provide excellent selectivity and reject spurious frequencies up to . The proposed SICL based filter is manufactured using low cost PCB fabrication process. Experimental results indicate an insertion loss of 1.18 dB and a 3 dB bandwidth of 360 MHz centred at 13.83 GHz. The self packaged SICL filter occupying only lateral area is suitable for implementation in densely integrated systems.

In this Letter, a new dual-band filtering power divider (FPD) with high isolation and high frequency selectivity by using self-packaged substrate integrated suspended line (SISL) technology has been developed. Both passbands of the proposed FPD can be designed and controlled individually. A pair of transmission zeros at either passband which are due to the multiple coupling paths with electric/magnetic couplings have been realised. The dual-band FPD has been fabricated layer by layer and then assembled, and the design has been demonstrated by experiment. The self-packaged SISL dual-band FPD has a circuit size of 0.41λ_g × 0.25λ_g.

A self-diplexing mixed-coupling bandpass filter based on a single slotted substrate-integrated-waveguide (SIW) cavity is proposed. By etching a wide slot, two unequal half-mode SIW parts are constructed and then fed by four coaxial ports to realise two bandpass filters with different centre frequencies. Since this etched slot can serve as a divider and an isolator, a flexible frequency ratio of 0.60–1.66 and high isolation of better than 20.0 dB is achieved. By further cutting the other two narrow slots, four quarter-mode sub-cavities are obtained where the mixed coupling is induced in the pass bands and a transmission zero (TZ) is introduced at the stopbands. The coupling and TZ are controllable by simply adjusting the length of these cut slots, which makes the proposed design easily redesigned for practical applications. A prototype with a size of 0.59λ₀ × 0.59λ₀ × 0.03λ₀ has been fabricated. The measured centre frequency, TZ, 3-dB bandwidth, and insertion loss of the lower pass band is 2.51 GHz, 2.96 GHz, 340 MHz, and 1.67 dB, respectively, whereas those of the upper pass band is 3.06 GHz, 2.53 GHz, 370 MHz, and 2.44 dB, respectively.

The accuracy of contact resistance values of two-dimensional (2D) field-effect transistors extracted with the Y-function considering the impact of the intrinsic mobility degradation is evaluated here. The difference between methodologies that take this factor into account and ignore it is pointed out by a detailed analysis of the approximations of the transport model used for each extraction. In contrast to the oftenly used approach where the intrinsic mobility degradation is neglected, a Y-function-based method considering a more complete transport model yields contact resistance values similar to reference values obtained by other intricate approaches. The latter values are more suitable also to describe experimental data of 2D devices of different technologies. The intrinsic mobility degradation factor of 2D transistors is experimentally characterised for the first time and its impact on the device performance is described and evaluated.

In this Letter, a recorded single channel of 6.915-Gbit/s white-light visible light communication (VLC) system is reported and experimentally demonstrated. By optimising the bit-loading algorithm onto the direct-current optical OFDM signal and without using an optical blue filter, the high data rate can be achieved. After a free-space propagation distance of 1.5 m, the white-light beam diameter of 14 cm and illuminance of 795 lux are measured. The proposed white-light VLC system can provide both lighting and communication simultaneously with functional propagation distance.

The authors experimentally demonstrate an ultra-low repetition rate passive mode-locked fibre ring laser with a semiconductor optical amplifier works like as a gain medium and non-linear polarisation rotation. A unidirectional erbium-doped fibre amplifier provides additional gain to increase cavity power. By exploiting non-linear polarisation rotation combined with dispersion management stable 5.18 ps optical pulses with 67.2 kHz repetition rate and energy of 1.75 nJ were generated with an extinction ratio >35 dB. The central wavelength is 1557.89 nm and the optical 3-dB bandwidth 1.82 nm very appropriate for optical applications that often require high-peak-power, low repetition rate and picosecond pulses.

A novel topology of the ac–dc power converter to supply neutral-point-clamped inverters used in ac motor drives and wind offshore power grids is proposed. It is based on the T-type multi-state switching cell, which allows obtaining a five-level voltage waveform downstream of the input filter inductor. The converter presents a high voltage gain due to the transformer secondary windings associated with series-connected diode rectifiers. Prominent main characteristics of this topology include power factor correction, operation of the filter inductor at twice the switching frequency, reduced conduction losses in the semiconductors, and voltage balance across the output capacitors. To verify the operating principle of the converter a design example is described, considering that the converter specifications are: output power of 3 kW; rms input voltage of 110 V; total output voltage of 1 kV, which is equally divided among four capacitors and switching frequency of 20 kHz. Simulation results are shown to validate the theoretical assumptions and verify the behaviour of the rectifier in terms of dynamic response while also supplying unbalanced loads.

In this Letter, a demodulation based pre-filtering technique is proposed for fast detection of the single-phase grid variables. The estimation of grid voltage parameters is achieved through a Type-I synchronous reference frame phase-locked loop (PLL). The modified bandpass filter in the pre-filtering stage ensure a better rejection of the DC-offset and harmonics. Thus, the grid variables are estimated rapidly with the help of a simple proportional controller. Consequently, in-loop filtering burden and the effort required to tune the PLL is minimal. However, steady errors are witnessed in the fundamental amplitude and the phase information which are mitigated by a curve fitting technique. Importantly, the proposed PLL scheme has improved the dynamic performance under off-nominal frequency conditions with immunity against grid disturbances. The robustness of the proposed algorithm is experimentally verified vis-à-vis existing literature.

Due to the sparse nature of inverse synthetic aperture radar (ISAR) images, sparsity-driven algorithms have been developed for high-resolution ISAR imaging. Recently, alternating direction method of multipliers (ADMMs) has been introduced for sparse signal recovery. In this Letter, based on ADMM, a fast algorithm for ISAR imaging is proposed which has lower computational complexity than that of conventional sparse representation-based methods and achieves good robustness in low-signal-to-noise ratio conditions. Experimental results based on both simulations and real data of the Yak-42 airplane show the superiority of the proposed method.

With the development of the array signal processing techniques, more and more robust methods are proposed for non-stationary interference cancellation at the beamforming stage. Among these methods, null widening methods or well-known covariance matrix taper (CMT) methods are the research hotspots. However, the typical CMT-class methods modified the whole covariance matrix with the same taper matrix, which means that all the nulls produced by the covariance matrix will be widened to the same width in cosine angular domain. This will make a waste of degrees of freedom (DOFs) in the scene that different interference sources have a different degree of non-stationarity. In this Letter, also adopting CMT structure, a flexible widening method is discussed, which utilises the relationship between each eigenvector and the steering vector of each interference source. The proposed method can widen different nulls with different width, which can save the consumption of DOFs as many as possible according to the demand. Besides, as a subspace-based method, it can preserve the minimal sample support property. The simulation results show the effectiveness of the proposed method.

This Letter demonstrates the technical capability of millimetre wave pulse-Doppler radar to identify safety-critical operating status information for parked cars. A simple 94-GHz pulsed linear frequency modulated radar with stretch processing on receive provides range-selective detection of small-scale vibration along the length of a parked car. The resulting range-Doppler signature clearly differentiates whether the car's engine is running or not as well as whether the car's transmission is shifting between ‘park’ and ‘drive’. The rotational motion of an opening car door also produces a distinct signature in range-Doppler space. Incorporating this knowledge into future automotive sensors can alert oncoming vehicles to unanticipated actions of parked cars to avert potential damage or injury.

When a small number of snapshots are used, the performance of the direction-of-arrival (DOA) estimation method based on the least squares (LS) degrades severely because of inadequate estimation of the covariance matrix. Although the subspace-based DOA estimation methods were proposed to improve the performance of DOA estimation method based on the LS; however these methods are computationally complex, especially for a large number of array elements. In this Letter, the DOA estimation method based on the LS is improved by reconstructing the covariance matrix with diagonal loading, where the diagonal loading factor is computed automatically by estimating the signal power. The reciprocal of the array pattern is taken to calculate the spatial spectrum, where the peak values correspond to the estimated DOAs of signals. The proposed method can achieve better performance with few snapshots and low computational complexity. The effectiveness of the proposed method is verified by the numerical simulations.

Jitter and shimmer appear as symptoms of disordered voices and they are defined as the cycle-to-cycle variation of the cycle length and cycle amplitude, respectively. The frequency modulation of the phonatory excitation, i.e. vocal jitter, causes amplitude modulation (shimmer) of the speech signal due to the action of the vocal tract transfer function. In this Letter, the authors investigate the extent to which the vocal tract filtering affects the accuracy of measurement of vocal cycle perturbations of the phonatory excitation. To achieve this goal, the direct digital synthesiser is used to generate a series of glottal excitation cycles with the desired amount of vocal perturbations that crosses the vocal tract model to produce a set of 231 stimuli of sustained vowels [a]. Praat is used to measure the amount of perturbations. Results show that the values of the input shimmer inserted in the glottal excitation are smaller than the values measured by Praat. An undesired amount of shimmer is generated in the speech signal when the glottal excitation signal perturbed with jitter crosses through the vocal tract.

In this Letter, a new construction of complementary sequence sets (CSSs) is proposed by concatenating Golay complementary pairs (GCPs) which could have different lengths. The constructed CSSs have flexible sequence lengths, i.e. , where L and N are the lengths of two constituting GCPs. For example, the proposed method can be used to construct quadriphase CSSs of all lengths up to 100, except 87 and 95. The feasibility for practical applications of CSSs will be enhanced due to their flexible lengths and good peak-to-average power ratio property.