In this Letter, the aperture efficiency of a folded reflectarray is improved using a proposed element. This element is designed with a simple structure using a rectangle and split-ring combined structure to obtain a more accurate reflection phase for the folded reflectarray than that of a conventional element in the millimetre wave frequency. Thus, a high aperture efficiency of 55.04% is obtained at 77 GHz. The element design concept and procedure are presented in detail.

In Internet-of-things devices, to alleviate the performance degradation of the ground radiation antenna due to the large battery, two compact electromagnetic bandgap (EBG) unit cells are introduced between the ground radiation antenna and the battery. The ground radiation antenna is comprised a small loop-type circuit that is designed within a 5-mm × 5-mm ground clearance for 2.4-GHz WLAN applications. The compact EBG unit cell is a modified square split-ring resonator that is designed by meandering-line technology. The measurement results showed that the 3:1 VSWR impedance bandwidth of the ground radiation antenna increased from 30 to 90 MHz, and the average total efficiency increased from 19% to over 54% with and without the introduction of the EBG structure, respectively.

This Letter presents the design of a reduced aperture low sidelobe patch array whose aperture is much smaller than that of a conventional patch array since the spacing between adjacent elements is less than . Low sidelobes are achieved without implementing an aperture distribution. This results in more efficient patch elements of the array and therefore the overall gain is higher than that of a conventional array of the same aperture. Moreover, as patches are series-fed, the width of feed lines is not limited by patch impedances and no matching transformer is required in the design. The experiment shows that a gain of 15.3 dB at 16.5 GHz and a sidelobe level of −16.7 dB in the E-plane and −15.1 dB in the H-plane are achieved.

A 185–220 GHz four-stage differential wideband amplifier is presented. The bandwidth (BW) is significantly expended by using the coupled transmission line (CTL) technique and the staggered inter-stage wideband matching technique. CTL is adopted due to its high self-resonant frequency, high simulation accuracy, compact layout and wide BW characteristics. The amplifier achieves a peak gain of 10.1 and 3 dB BW of 35 GHz ranging from 185 to 220 GHz. With a supply voltage of 0.9 V, the amplifier consumes a DC power of 54.5 mW. The amplifier occupies a core area of 500 × 250 μm² in 40 nm CMOS.

A highly linear folded-type up-conversion mixer working at 10 GHz is presented. The mixer utilises a feedback loop to linearise the g_m stage and make the effective g_m only determined by a resistor theoretically. Moreover, the effects of local oscillator (LO) power levels on the circuit DC condition and linearity performances are examined from both the theoretical analysis and measurement results. The mixer prototype, fabricated in a 65 nm CMOS technology, consumes 7.7 mW from 1.2 V supply voltage and exhibits 13.4 dBm OIP3 at the LO power of 2 dBm.

This Letter studies the non-ideal analogue filters performance, time delay and phase offset in hybrid filter bank (HFB) DACs. After compensating the non-ideal filters performance using a digital signal, a linear phase function containing time delay and phase was deduced based on the auto-power spectrum of the HFB DAC output in the overlapped frequency band. ‘Three-point’ method was used to simultaneously estimate these two errors by unwrapping the phase affected the principal value interval of the inverse trigonometric function. These two errors were compensated by the digital pre-distortion technique. Finally, simulation results were provided to verify the effectiveness of this proposed method.

By introducing a tiny perturbation into a memristive chaotic oscillator, a new memristive chaotic system without equilibrium is proposed in this letter. Therefore, the resulting attractors are all hidden. Particularly, the attractive phenomenon of hidden extreme multistability is observed. The complex dynamical behaviours are numerically analysed. Furthermore, a hardware experiment is carried out. A good similarity between numerical simulations and experimental results is obtained, which verifies the feasibility and effectiveness of this method.

In the security system, there is a lot of sensitive information, and the leakage of the secret key will cause security failure. In this Letter, combining the fault injection analysis of SM4 (Chinese block cipher algorithm standard), a key leakage Trojan payload circuit that requires only 4 XOR gates was designed and verified. By estimation, the area and power consumption is <1% of that in popular solutions. The scheme was successfully verified using FPGA. This Letter makes contribution to the defence for key-leakage type Trojan.

Meshed ground planes have been used in a flexible printed circuit board (FPCB) to raise the characteristic impedance of a stripline to 50 Ω. However, the characteristic impedance varies considerably with its position on the meshes. An offset mesh structure is proposed to address this issue. Both simulation and measurement results verify its effectiveness.

The performance of open-loop ADCs implemented with VCOs is limited by VCO non-linearity and first-order noise shaping. The resolution limitation imposed by first-order noise shaping can be compensated by a ring oscillator VCO with many output phases. Linearity can also be improved by using a feedback loop around the VCO closed with a DAC. However, a long ring oscillator may require a DAC with a prohibitive number of bits if feedback is used to compensate distortion. This Letter proposes an ADC architecture based on the Leslie-Singh Sigma-Delta () modulator that allows to implement a distortion correction loop around a VCO with a simplified DAC of few levels, yet keeping a large number of output quantisation levels to maintain resolution. The Letter discusses the system-level architecture and shows an implementation circuit example to verify the effective correction of distortion.

This Letter deals with the tuning of series proportional–integral–derivative (PID) controller to provide desired time-domain specifications. A simple analytical design method based on pole placement is developed. First, the integral and derivative time constants are obtained. Then, a good performance–robustness trade-off is achieved by suitably choosing the controller gain in the stability range. To evaluate the performance of the proposed method, an illustrative example is presented. Moreover, the simulation results are compared with those given by Skogestad internal model control, a well-known PID tuning method.

A new active contour model driven by harmonic mean based Kullback–Leibler (KL) fitting energies is presented to achieve accurate segmentation of various kinds of images. Firstly, the global fitted image is defined by global harmonic mean region fitting centres and the piece-wise smooth function and then the global fitting term can be constructed based on KL divergence. Secondly, the local fitted image is defined by local harmonic mean region fitting centres and the piece-wise smooth function and then the local fitting term is also constructed based on KL divergence. Moreover, the regularising term is incorporated into the energy functional. Experimental results of real-world images verify the effectiveness and superiority of the presented model.

The location of the eye is an important feature for computer vision and pattern recognition applications including psychological analysis, facial expression recognition and auxiliary driving. Most of the current work typically performs face detection first, followed by eye centre localisation. Such an algorithm fails to locate eye centre in some cases because the illuminance and external environment changes the facial features blurred in the images; therefore, the traditional face detectors cannot accurately determine the region of the face. A method for eye centre localisation which is based on convolution neural networks and regression framework is proposed. First, eye region is selected from the facial images according to convolutional neural networks. Then, the accurate eye centre is estimated by a linear regression model. Finally, the method on BioID and GI4E datasets is evaluated. The experimental result showed that the proposed algorithm has comparable or better performance compared with the state-of-the-art methods.

A novel and efficient metric learning strategy for person re-identification is proposed. Person re-identification is formulated as a multi-domain learning problem. The assumption that the feature distributions from different camera views are the same is overthrown in this Letter. ID-based transfer component analysis (IDB-TCA) is proposed to learn a shared subspace, in which the differences in the feature distribution between source domain and target domain are significantly reduced. Experimental evaluation on the CUHK01 dataset demonstrates that metric learning with IDB-TCA embedded outperforms state-of-art metric methods for person re-identification.

Spectral clustering (SC) is a kind of clustering method which shows superior performance in recent years; however, the method is sensitive to the initial scale parameter. In order to improve the accuracy and stability of clustering, this Letter proposes a novel approach of SC ensemble model based on distance decision. First of all, the model performs several times clustering for data samples by SC, and corresponding joint Laplacian matrix is obtained. Then, a new method of distance decision is proposed, the joint Laplacian matrix is transformed into a cumulative distance matrix using the fuzzy theory. Finally, the distance matrix is introduced into the density peaks (DP) algorithm, and the final results of clustering are obtained by the improved DP algorithm for clustering ensemble. The experimental results show that the clustering ensemble model proposed in this Letter is more effective than other classical clustering ensemble model on the 12 data sets in University of California Irvine machine learning database.

A polar code-based overlapping transmission scheme to improve channel utilisation for Internet of Vehicles is presented. In the presented scheme, different relay nodes (RNs) are allowed to send their packets to the destination node (DN) in the same time slot and frequency band. The signals from RNs to DN are overlapped and hence non-orthogonal. However, the information in the overlapped signals can be recovered by the proposed decoding algorithm. Numerical results show that the decoding algorithm performs well on information recovery in the presented transmission scheme.

In this letter, the authors introduce an adjustable window function that combines the ability of ultra-spherical window function to control peak side-lobe level and asymptotic side-lobe attenuation, with the side-lobe pattern of the Taylor–Villeneuve window function where side-lobes near the main-lobe have constant amplitude, which provides additional advantage over the ultra-spherical window function of lower main-lobe width. They proposed frequency and time domain window function designs that are obtained by modifying the properties of the power of Hann window function. In both cases, the asymptotic side-lobe attenuation is given in closed form as a function of a single parameter.

In recent years, two new types of irreducible pentanomials, i.e. Type C.1 and Type C.2 pentanomials, and their associated generalised polynomials bases (GPBs) have been proposed to yield efficient bit-parallel multiplier architectures. The GPB squarer for Type C.1 pentanomial is also investigated previously. But no GPB squarer for Type C.2 pentanomial is given as these pentanomials are far more complicated. In this Letter, the authors give explicit GPB squarer formulae for all Type C.2 pentanomials by re-classifying these pentanomials into certain sub-groups, which is based on the parities of pentanomial parameters. As the main contribution of this Letter, the authors show that the GPB squarers for most Type C.2 pentanomials match the fastest results.

Minimum average correlation energy (MACE) filters are initially developed and widely used for image pattern recognition tasks. A novel method leveraging the enhanced MACE filters is proposed to tackle classification problems from a new perspective. By employing 1D Fourier transform, establishing new identification metric, and improving numerical stability, the proposed method constructs an enhanced correlation filter to select a sub-population of the un-labelled data, and subsequently outputs labels. Experiments show our method achieves 100% precision on multiple datasets considered, including two public benchmarks and one obtained from semiconductor industry addressing an emerging task.

A new miniaturised dual-mode balanced bandpass filter (BPF) is presented in this letter. To achieve the required differential-mode (DM) operation with common-mode (CM) suppression, an octagonal microstrip dual-mode loop resonator is employed. By symmetrically locating four high-low impedance sections inside the free area of the loop resonator, the proposed balanced BPF can realise compact size, wide stopband under DM operation, and extended CM suppression, simultaneously. Furthermore, a pair of coupled lines is introduced to readily achieve two transmission zeroes for realising high selectivity of DM response. For demonstration, a dual-mode balanced BPF with the passband centred at 2.39 GHz is designed and measured. Both the simulated and measured results are provided with good agreement.

A broadband and compact rectangular waveguide twist are presented and verified. Adopting an axially rotating 45° ridge waveguide, the rectangular waveguide twist provides a very compact interconnection of two rectangular waveguides with different on-axis angle rotation. Compared to conventional waveguide twist, the superiority of the proposed waveguide twist is the easiness for manufacturing and the overall dimension is compact, the thickness is only 0.25 times the waveguide wavelength of the TE₁₀ mode. For verification purpose, a Ka-band 90° waveguide twist has been designed, fabricated, and tested. Measurement results show that the return loss is <−20 dB and the insertion loss is in excess of −0.2 dB. The measurements are in good agreement with simulations and demonstrate the excellent performance of the waveguide twist at the Ka-band.

In this letter, a quintuple-mode bandpass filter for a UWB system with improved performance is presented. The presented filter is implemented with asymmetric parallel-coupled lines loaded open stub and short stub connected at the symmetrical plane of the filter to increase its frequency selectivity. Five resonant frequencies within the UWB band and two transmission zeros at the lower and upper passband/stopband edges are observed. A filter prototype is fabricated for verification and the experimental data are corresponding well with the simulated results. The proposed filter achieves a bandwidth that can exactly cover the whole UWB frequency band. Moreover, this filter has a compact size (0.645λ _g × 0.318λ _g) and low insertion loss.

The development of compact amplifiers with an average power of the frequency range 0.2–0.3 THz is related to the design of electron-optical systems using converging sheet beams that allow obtaining sufficiently high current densities with a lower current load on the cathode and with a lower value of the magnetic field. Based on the theory of synthesis the numerical simulation results for electron-optical systems for the formation of a converging sheet electron beam are presented. Reduction of the current load on the cathode allows the principle of transformation of a cylindrical electron beam into a sheet by employing electrostatic compression of the beam in one plane. A sheet beam electron-optical system with a field emission cathode with convergence of 10, and with the beam of 0.1 × 1.1 mm² and current density of 45 A/cm² has been performed. The electron gun employing the cathode–gate structure with diameter of 1 mm has an emitting current up to 10 mA in experiment. The modulation coefficient of the cathode current for a gun with a sheet electron beam has 0.3% of the anode potential.

Towards the realisation of reliable communication systems using simple nanoscale transceivers, this Letter presents a method for the non-coherent detection of digitally phase-modulated signals. Without the synchronisation of the phases of the transmitted and received carrier signals, data transmission cannot be reliably achieved. To address this issue, the proposed method is developed using a nanoscale receiver that receives the binary phase-modulated signal through the mechanical vibration of a carbon nanotube cantilever. The present analysis reveals that this receiver has an interesting characteristic: the two decision statistics available in the receiver have a phase difference of , which means that at least one of the statistics is fairly reliable regardless of the synchronisation error. Through the selection of the more reliable of the two statistics, the proposed method achieves low error in the estimation of the transmitted data even without carrier phase synchronisation. This Letter presents the derivation of the proposed rule for selecting the statistic and numerical results that demonstrate the effectiveness of the proposed method. This Letter is expected to contribute to the realisation of future communication applications using nanoscale transceivers.

Based on electrical frequency sweeping, an ultrahigh-resolution coherent optical spectrum analyser with a doubled measurement range is proposed and experimentally demonstrated. In the scheme, a frequency-swept electrical signal is converted to the optical domain. By selecting one of the second-order sidebands of the generated optical signal, a wavelength-swept signal with a frequency range that is twice the bandwidths of the electro-optic and RF components is generated. With the wavelength-swept signal, high-resolution optical spectrum analysis with a doubled measurement range is implemented. In a proof-of-concept experiment, the spectrum of a multi-tone optical signal with a bandwidth of 20 GHz is measured with a resolution of 1.5 MHz by employing 10 GHz components.

Owing to unstable multi-wavelength lasing, robust measurement is difficult with a fibre Bragg grating (FBG) laser sensor system that uses FBGs as laser cavity mirrors as well as sensors. A stabilising technique is essential for proper data acquisition. Here, it is demonstrated that polarisation scrambling (PS) stabilises the output of laser sensor system in combination with a spectrometer demodulation. The state of polarisation (SOP) of a fibre-optic laser cavity is changed by a 1 MHz PS. A 5 kHz spectrometer traces the distribution of output laser wavelengths. The rapid changes in SOP make all the FBG wavelengths lase as long as necessary to determine the Bragg wavelength shifts in the sampling windows of a photo-diode array. A laser sensor system with eight FBG sensors is used to demonstrate robust measurement and data collection under environmental perturbations.

An FPGA-based parameter estimation method for linear frequency modulation (LFM) signal is proposed. The average frequency value of basic time slice is obtained by using improved time-domain counting method. Then, frequency values are used for the least-squares fitting, so as to obtain the sweep rate of LFM signal. Compared to estimation methods based on digital signal processing, the proposed approach can effectively reduce the estimation time. Experimental results show that the slope fitting accuracy of mono-component LFM signal is better than 99.8% and the processing time is <250 ns.

Normally off p-gallium nitride (GaN) gate high-electron-mobility transistors (HEMTs) on silicon substrate were fabricated with hydrogen plasma treatment technology, which features a high-resistivity cap layer (HRCL) at the access region. With hydrogen plasma treatment at the access region, the normally off operation was obtained with a threshold voltage of +2.5 V based on the linear extrapolation of the transfer curve. The fabricated HRCL-HEMT with a gate width of 49.7 mm exhibits a maximum drain current of 10 A at V _GS = 8 V and a high off-state breakdown voltage of 567 V at V _GS = 0 V with substrate grounded. Meanwhile, the HRCL-HEMT also shows low-gate leakage current of 1.3 × 10⁻⁷ and 3.3 × 10⁻¹⁰ A/mm at V _GS = 8 V and = −30 V, respectively. The positive threshold voltage, high drain current, high-breakdown voltage and low-gate leakage show the potential of HRCL-HEMT for power switching applications.

Indoor localisation has attracted a lot of attention because of its importance for location-based services. A fusion algorithm (named as YELM-DS) based on extreme learning machine (ELM) and Dempster–Shafer (D–S) evidence theory is proposed. ELM learns the input data model composed of inertial and visual information and target output positions with high speed. During online phase, the final localisation result of a frame is decided by the trust degree obtained from D–S. Angle judgments are also introduced to decrease the big localisation errors of turning. Compared with the existing vision-only methods, the proposed method can both run in real time and achieve good localisation accuracy even in challenging scenarios.

The effective capacity of a communications system using unified models is analysed. To obtain a simple closed-form mathematically tractable expression, two different unified approximate models have been used. The mixture gamma distribution which is a highly accurate approximation approach has been first employed to represent the signal-to-noise ratio of fading channel. In the second approach, the mixture of Gaussian distribution which is another unified representation approach has been utilised. A comparison between the simulated and numerical results using both distributions over composite /gamma fading channels have been provided.