Industrial monitoring and process control, power plants, aerospace industry, military equipment manufacturing, oil and gas industries are examples of businesses in need for high-temperature and harsh-environment electronic components and systems. In particular, resonators and filters that operate beyond the 125°C military range upper limit, normally dictated by silicon-based semiconductor devices, are required in applications which demand frequency control, clocking and sensors. In this Letter, temperature compensated surface acoustic wave orientations appropriate for the fabrication of resonators and filters >125°C have been identified through numerical calculations and experimentally confirmed on a commercially available langasite wafer, a piezoelectric crystal which operates at high temperatures. Resonator filters have been designed, fabricated and tested along two orientations, confirming the zero temperature sensitivities at 150 and 300°C. These devices are of great interest for modern harsh environment frequency control, timing and sensor applications.

This Letter proposes a circularly polarised (CP) monopole antenna with filtering function, which is achieved by integrating a miniaturised coupled filter based on the bent-loop resonators. The circular-polarisation is obtained by using the falcate-shaped patch as the radiation part, and the existence of the coupled filter in the feeding line would not affect the CP radiation performances. Both numerical simulation and experimental demonstration are performed for the designed CP filtering antenna, and good agreements are obtained. The designed antenna with dimension of 70 × 75 mm² exhibits good frequency selective performances for the return loss, gain and axial ratio at 2.45 GHz with measured impedance matching bandwidth (|S ₁₁| < 10 dB) of 0.23 GHz. Such CP filtering antenna is a good candidate for the emerging application area, e.g. wearable biotelemetric devices.

A new method for suppressing surface waves on a millimetre-wave microstrip array antenna is presented. A 10 × 2 patch array antenna is designed at 77 GHz for W-band radar applications. Complementary split ring resonator (CSRR) has been designed at 77 GHz, etched on the antenna ground plane. Simulation and experimental results are compared in terms of the gain and the side lobe level with and without CSRR. CSRR between series patch antenna array could suppress surface waves effectively so that this approach enhances the gain and reduces the side lobe level. The measured gain improvement in the antenna with CSRR is 2.5 dB. The side lobe level of the antenna with CSRR is reduced up to 3.5 dB lower than antenna without CSRR.

An air-filled substrate integrated waveguide (AFSIW) slot array antenna based on multilayer PCB process is proposed. For comparison and demonstration purposes, 1 × 4 slot array antennas based on both AFSIW and conventional dielectric-filled SIW (DFSIW) operating at 30.5 GHz are designed, fabricated and measured. The exampled AFSIW 1 × 4 slot array antenna achieves 44 and 18% narrower measured H-plane and E-plane half-power beamwidth, respectively, and simulated radiation efficiency improved by 1.6%, resulting in an overall measured gain improvement of 3.5 dB compared with its DFSIW counterpart. This antenna is of high interest for the design of efficient arrays based on high-performance AFSIW feeding network.

A low-noise, wide-bandwidth DNA readout instrument for nanopore applications is presented. Owing to hardware simplicity and reliability, a resistive-feedback transimpedance amplifier (rf-TIA) is adopted as the headstage for the readout instrument. However, to achieve a high gain and low input noise, its high feedback resistance induces a high parasitic capacitance, thus significantly limiting the 3 dB bandwidth. To drastically reduce the parasitic capacitance and widen the bandwidth, a novel rf-TIA architecture is fabricated that utilises a split-resistor technique for the high feedback resistor. This is demonstrated in a benchtop test employing an α-haemolysin nanopore.

A high-gain and high-speed gain-boosted dynamic amplifier for pipelined-successive approximation register ADCs is presented. A cross-coupling cascode topology is proposed to boost the gain of a single-stage dynamic amplifier. The proposed dynamic amplifier achieves a voltage gain of 32 under a 1.2 V power supply in 130 nm CMOS technology. Sampling at 150 MS/s, the proposed dynamic amplifier consumes a total power of 0.22 mW.

Four types of liquid-metal-based tunable capacitors with a high tuning range are demonstrated. The capacitors are easily fabricated by placing a pair of liquid-metal-filled tubes in contact with one another. Tunability is achieved by varying the length of the liquid metal in one of the tubes using a mechanical pump. Four different structures are demonstrated: parallel-tube, folded-tube, coil, and spiral capacitors. The highest measured tuning ratio is 42:1, and the highest change in capacitance per unit length of the pumped liquid metal is 0.07 pF/mm.

A power efficient noise-coupled ΔΣ analogue-to-digital converter (ADC) is proposed. The new architecture can effectively implement double-noise coupling with simple digital delay cells. For performance improvement, one more input feed-forward path is added to reduce the output swing of an integrator. Zero optimisation technique was also applied to the proposed modulator to increase the signal to quantisation noise ratio (SQNR). A third-order ΔΣ ADC using new configuration was designed and simulated. The results verify the advantages of the proposed architecture.

A discrete-time ΔΣ analogue-to-digital converter utilising a multi-bit successive approximation register quantiser and a segmented digital-to-analogue converter (DAC) is proposed. By adding the delayed DAC output to the modulator input and removing it at the modulator output in the digital domain, the mismatch error is filtered. Thus, a first-order DAC mismatch error shaping is realised. This improves the DAC linearity without introducing an extra delay in the feedback path. Segmenting the DAC and processing the least significant bit and most significant bit segments separately, the dynamic range remains unaffected.

Using Gaussian mixture model (GMM) as a patch prior has achieved great success in image restoration (IR). However, based on the derivations, the objective function of the conventional GMM-based IR methods can be expressed as the minimum of mean square error (MMSE) criterion based optimisation problem. As outliers may exist in the patch cluster obtained even within a spatially constrained window and MMSE is sensitive to outliers, using the MMSE criterion based strategy would lead to error results. A maximum correntropy criterion (MCC) based IR method with GMM as a patch prior is proposed. The proposed method can automatically identify outliers and assign weight for each patch within the k-nearest-neighbour patches with respect to each exemplar patch in the image and thus can robustly estimate the Gaussian parameters which results in more accurate estimation of the image patches. Finally, an effective iterative optimisation algorithm is designed to solve the proposed objective function under the MCC criterion. The experimental results for image inpainting demonstrate the capabilities of the proposed method.

The problem of image compressive sensing (CS) preprocessing is considered. Currently, image CS reconstruction algorithms mainly consider the sparsity prior knowledge of original image. However, the change of the sparsity strength among the different images may degrade the efficiency of the reconstruction algorithms. Thus the idea of CS preprocessing is proposed to serve two purposes: strengthen the sparsity property of the CS measured image and make preprocessing and reconstruction algorithm matched. Specifically, the collaboration reduced rank (CRR) preprocessing is proposed based on non-local sparsity and non-local low-rank regularisation reconstruction algorithm (NLR-CS). Then a more efficient CRR-NLR-CS CS reconstruction method is proposed which utilises the CRR preprocessing and NLR-CS. Experimental results show the effectiveness of the proposed method.

A novel method of facial expression recognition using graph signal processing (GSP) is proposed. Here the dimensionality of the feature vector of the facial expression based on the histogram of oriented gradients and discrete wavelet transform is reduced using the GSP and then applied to a classifier. The proposed scheme is compared with the existing methods on the JAFFE dataset in terms of the accuracy and the length of the feature vector. It is seen from the simulation results that the proposed scheme provides better accuracy than the existing methods.

The bit error rate (BER) performance of 2 × 2 full-rate linear-receiver space–time block code (FRLR-STBC) is evaluated. By using the Appell and Gauss hypergeometric functions, the novel exact and asymptotic closed-form BER formulas of FRLR-STBC with BPSK modulation, along with the corresponding coding gain and diversity order, are derived over Rayleigh fading channels. Some numerical results are presented to demonstrate the validity of the derived analytical results.

As an essential function in software defined networks (SDN), the link failure detection typically runs at all SDN controllers on a periodical basis. To reduce the number of messages processed by the SDN controller, an efficient and lightweight link failure detection mechanism is proposed, where switches actively inform SDN controller about the faulty link. Hence, SDN controllers do not need to process feedback from switches in each detection round. The experimental results show that compared with the most commonly-used failure detection mechanism, the proposed one can significantly reduce the overhead of the SDN controller.

‘Internet of things’ (IoT) operates for all-time availability of network components to provide a wide range of service applications to the network users or clients. IoT operates for a large number of applications, and one of these include the sensor deployment for data support to clients. However, these sensors require a large number of queries to be handled by the network components considering the constraints of memory, energy, and delays. These issues can be easily tackled by deploying unmanned aerial vehicles (UAVs), which serve as dynamic nodes to handle clients’ queries by acting as on-demand gateways. However, this UAV-assisted IoT requires efficient localisation to determine the sensor for handling multiple-content queries, which is considered as a problem. A cost function is formed in the proposed approach to generate an optimisation problem which is resolved using a radial basis function kernel support vector machine. The effectiveness of the proposed approach is demonstrated in terms of significant gains attained in terms of error in the cost function, overheads in handling queries, and accuracy in UAV allocation.

A balanced ring hybrid with arbitrary power division ratio is proposed. By embedding four half-wavelength microstrip lines in the traditional single-ended ring hybrid, the functions of differential-mode power division, common-mode suppression, arbitrary power division ratio and port isolation can be achieved at the same time. Higher power division ratio also can be obtained compared with traditional single-ended design for the achievable maximum characteristic impedance of 120 Ω. A simplified analysis process is proposed by making the balanced port equivalent to a new single-ended port. To verify the theoretical prediction, one prototype with a power division ratio of 4 : 1 and a size of 0.59 × 0.56λ _g (λ _g is the guided wavelength at the centre frequency) is measured and compared with the reported designs. The measured results exhibit the minimum insertion loss of ∼0.2 dB and the operating bandwidth for |S_{dd AA}| < −15 dB of 20.9%.

An n-pole absorptive reciprocal filter consists of at least 2n single-mode resonators is proved. An absorptive band-pass filter with four-pole Chebyshev response is designed with two four-pole filters and two-quadrature hybrids. The quadrature hybrids are used to divide the signals into two paths. Additional two ports are introduced to absorb the reflected energy. This circuit structure is easy in fabrication on superconducting wafers. The filter is successfully designed and fabricated on an MgO substrate. The measured results match well with the simulations.

Two four-order dual-mode rectangular waveguide filters centred at 100 GHZ with return loss all better than −19 dB are presented. The filters consist of two cascade resonant cavities suppose TE102/TE301 degenerate modes and three pairs of round corner inductive irises symmetrical with waveguide. Since the introduction of round corner irises and dual-mode cavities, the filter structures have lower aspect ratio and less resonant cavities, thus the difficulty of fabrication sharply reduced. Single zero exists at the one side of pass-band because of the same width of two resonant cavities. All the other parameters constant except for the widths of two cavities slightly unequal, two distinct transmission zeros can generate at the same side of pass-band. UV lithography technology with SU-8 photo-resist is used for the fabrication of filters. On the whole, results from measurement are in good agreement with simulation, which fully confirm the validity of design and fabrication methods.

Variations of low-frequency noise including flicker and Lorentzian spectra are the dominant source of variation of noise equivalent power (NEP) for terahertz detectors using diode-connected N-type MOS (NMOS) transistors. The low-frequency noise of Schottky diodes fabricated in the same CMOS process as the NMOS detectors are dominated by generation and recombination noise. At 1 MHz modulation frequency, the noise of Schottky barrier diodes (SBDs) and its variation are ∼2× lower than those of diode-connected NMOS transistors. It should be possible to reduce the variation of NEP by ∼2× using Schottky diode detectors modulated at 1 MHz.

A microstrip planar lowpass filter (LPF) with sharp roll-off and ultra-wide stopband up to 40 GHz is presented. The central part of the LPF consists of three basic circular sectors, two smaller quadrants and a number of high-impedance lines that are cross-coupled to form a high cut-off rate and a moderate stop band from 2.3 to 10 GHz. Then, four types of resonator cells, three of which are high-impedance lines loaded with semi-circulars and one of which comprises folded step-impedance lines, are added to the central part to generate additional transmission zeroes up into Ka band. As a result, an ultra-wide stopband with 18th-harmonic suppression has been attained. A demonstration filter with 3 dB cut-off frequency of 1.96 GHz has been designed and fabricated. The measured results show that the relative stopband bandwidth of the LPF is 180% with suppression level >25 dB, and the roll-off rate is 104 dB/GHz. The functional area size of the LPF is 18.4 × 11.6 mm, which corresponds to 0.189 × 0.121λ _g (λ _g is the guided wavelength at f _c).

Optical 64-QAM transmitter using QPSK modulators in tandem driven by binary electrical signals is proposed. The proposed transmitter eliminates the need for complex electronics required for the generation of multiple-level signals and provides a Gray-coded constellation with high quality eye diagrams. The design could be scaled to higher-order QAM based on the user needs.

A microelectromechanical system capacitive pressure sensor with a differential configuration (patent pending), including a reference and sensitive capacitor, is presented. A single cavity includes both reference and sensitive electrodes. This implementation is achieved by a size reduction of the sensitive electrode. A reduction of the 40% on the sensible electrode area represents a decrease in the sensor sensitivity of only 12.5%, and allows the reference capacitor implementation in the area commonly used only for the sensitive capacitor. A tridimensional model of the sensor was simulated using a finite-element solver considering coupled mechanical and electrostatic models. The simulations show that this pressure sensor has a sensitivity of 4.2 fF/bar. The proposed pressure sensor designs use only materials available in complementary metal oxide silicon (CMOS) standard processes. A CMOS compatible post-process, based on the sacrificial material technique and its restrictions is presented.

High-density and high-speed charge-trapping AND flash memory array is fabricated for the first time. A reliability of 10⁴ endurance cycles and uniform program/erase characteristics along with a threshold voltage window >3 V is obtained. The AND array has several advantages, such as high read current drivability regardless of the number of word-lines, immunity to back-pattern dependency, and fast bit-sensing speed based on a parallel connected cell array structure, which are highly appropriate for three-dimensional (3D) stacking. Finally, a novel 3D stacked vertical-AND array is proposed to surpass the limitations of the conventional 3D NAND flash memories.

We demonstrate a simple polymer Topas-based low-loss hollow-core photonic crystal fibre for efficient guiding of terahertz radiation. The guiding properties are numerically investigated using finite-element method. It is shown that the proposed fibre exhibits extremely low average effective material loss of 0.017 cm⁻¹, very small near-zero average group velocity dispersion of 0.263 ps/THz/cm with low total loss over 2–2.20 THz bands. Moreover, high air-core power fraction ∼94% having single-mode guiding is also reported for the optimum design parameters.

In this work, single-mode distributed feedback interband cascade lasers with tapered ridges are presented. The lasers consists of a 500 µm long and 5.7 µm wide ridge wave guide section followed by a 1400 µm long tapered section with an output facet width of 35 µm. The device is operational in continuous wave mode and well above room temperature. At 25°C, the threshold current and threshold current density were 81 mA and 324 A/cm², respectively. The authors demonstrate a tuning range of over 14 nm with side mode suppression ratios >25 dB. The device showed temperature and injection power tuning coefficients of 0.35 nm/K and 4.09 nm/W, respectively. Output powers of 18 mW single mode and 35 mW multimode were achieved.

Design and fabrication of a monolithic, single aperture widely tuneable diode laser emitting around 970 nm is presented. Overgrown sampled grating Bragg reflectors are applied to setup wavelength selective mirrors. Wavelength tuning is achieved via the Vernier effect with thermal heaters placed on top of the sampled gratings. The 20 nm quasi-continuous tuning around 970 nm with an optical side mode suppression ratio in excess of 40 dB is demonstrated and an optical output power of more than 30 mW at a pump current of 180 mA is presented.

The adjustable speed range for three-phase four-switch inverter (TPFSI) fed permanent magnet synchronous motor (PMSM) drives is seriously limited by the DC-link capacitor voltage offset in TPFSI. For the traditional offset suppression schemes, the low-pass filters or notch filters are necessary to extract the capacitor voltage offset, which will deteriorate the control performance of the offset suppression schemes severely, especially for the low-speed range. Thus, a novel offset suppression scheme with no filters is proposed to improve the situation. For the proposed scheme, to extract the offset with no filters, a new static coordinate is defined, the relationships between the fundamental/offset components of capacitor voltages and the load currents are revealed, based on which a simple algorithm is developed. Then, the desired compensatory current is obtained by a closed-loop control and injected into the stator current control loop. Experimental results have validated the proposed scheme.

To improve the accuracy and robustness of high-resolution range profile (HRRP) target recognition, in this paper, the multi-scale fusion sparsity preserving projections (MSFSPP) approach is proposed for feature extraction. Compared with traditional multi-scale feature extraction method, the proposed MSFSPP approach utilises features in every scale and their sparse reconstructive relationship to construct multi-scale fusion features which contain more discriminating information. Support vector machine is employed to verify the classification performance of features extracted by MSFSPP and related feature extraction methods. Simulation results based on the measured aircraft datasets show that the proposed MSFSPP approach has outperformance with a small amount of data.

Radar measurement accuracy is mainly dependent on the signal-to-noise ratio (SNR). SNR variation of the same target is related to fluctuating target radar cross-section (RCS) that is known as Swerling RCS model. Radar measurement accuracy is an important parameter to design a radar system or to analyse its performance. More useful measurement accuracy formula than the previous research that is sensitive to the detection threshold is proposed. To investigate the measurement accuracy affected by Swerling model, the relative standard deviation is defined to compare the value of the previous research with the proposed value.

Outage probability (OP) of amplify-and-forward direct-sequence code-division multiple-access (DS-CDMA) systems is analysed with best path selection over α − η − μ fading channels. First, cumulative distribution function (CDF) of α − η − μ fading is derived and then, lower bound of the OP is obtained for the considered system. An asymptotic CDF expression is also proposed in order to get more insights on system behaviour. The OP performance is presented for different relay and fading parameter configurations. Analytical derivations are validated by the close agreement with simulations.