A two-port synchronous resonator was used for measurements of surface acoustic wave (SAW) parameters in the Y-plane of NdCa₄O(BO₃)₃ crystal. For YX–5°, YX and YX + 5° orientations, the SAW phase velocity under free surface, electromechanical coupling coefficient, second-order temperature coefficient of frequency and turnover temperature (T _O) of about 3481, 3502 and 3503 m/s; 0.68, 0.63 and 0.6%; −37, −32 and −70 ppb/°C² and 80, 25 and −35 °C, respectively, were obtained. The crystal was assessed as attractive in SAW applications due to temperature compensation and high electromechanical coupling coefficient.

The use of liquid metal as a switching mechanism for a frequency-reconfigurable edge-fed microstrip patch antenna designed to operate in either the 2.4 GHz ISM band or the 1.6 GHz GPS band is illustrated. A partitioned fluidic manifold bonded to the top of the antenna enables the pressure-driven displacement of a low-loss dielectric fluid and the liquid metal from a reservoir atop the patch to switch between the two antenna states. Fabrication and measurement of the antenna illustrates the use of fluidics as an alternative to RF switches for reconfigurable antennas.

A new approach to interfacing fabric based wireless systems to traditional electronics circuitry is presented. Measurements show possible insertion losses of the connector to be < 1 dB up to frequencies of about 4 GHz when electroplating has been used to increase the hook and loop conductivities. This reported work points towards an interconnection system that is flexible and will allow the removal and replacement of conventional electronics connected to textile based systems.

In the presence of steering vector model mismatches, a novel robust adaptive beamformer is proposed based on interference-plus-noise covariance matrix reconstruction. An interference region is first selected according to the coarsely estimated desired signal direction, and then the interference steering vectors are estimated with the robust Capon beamformer. Finally, the interference-plus-noise covariance matrix is reconstructed by integrating the Capon spatial spectrum over the interference region by using the estimated steering vectors.

A new approach for a multi-class steady-state visual-evoked potential (SSVEP)-based brain–computer interface (BCI) is proposed. It was demonstrated through preliminary experiments that spatial patterns of SSVEP responses recorded using high-density electroencephalography while presenting pattern reversal checkerboard stimuli with different spatial patterns can be classified with fairly high accuracy. The average classification accuracies in two of the three subjects were 91.7% (12-class) and 93.3% (15-class), suggesting that the proposed visual stimulation can potentially be used for multi-class SSVEP-based BCI.

To analyse and compare physical unclonable function (PUF) concepts, a set of parameters should be defined to measure performance. The Hamming distance (HD) is a parameter to measure the number of different elements of two strings of the same length. In this reported research, weighted HD is proposed to evaluate and compare the performance of PUFs. The proposed parameter uses weights for calculating similarity measure. The weights can eliminate the effects of noisy responses. The proposed parameter is more efficient and has a low overhead and can be used in identification and authentication applications.

A comparison of particle swarm optimisation (PSO) and market-based optimisation (MBO) is presented when applied to obstacle avoidance by mobile robots using artificial potential fields and special traffic rules. Most notably, PSO and MBO are applied to optimise the motion of mobile robots when acting in a common confined workspace. Simulation results show that both methods perform equally well with slight advantage for PSO.

A novel method for face recognition that accounts for both expression and lighting variables within a single model is proposed. It performs recognition by calculating a dense matching between the face images on a minimum intensity difference tree, which can be both extremely fast and accurate. Very promising results are presented on the AR Face Database when compared with other algorithms.

Since the data storage capacity per unit area of a QR code is very limited, it is restricted in various QR code applications. An approach is proposed based on visual skew to improve the data capacity of the QR code. The proposed method can enhance the capacity of the QR code per unit area by using different angle views to combine several QR codes together. The storage capacity of the proposed mechanism can be increased 1.5 times per unit area more than that of the normal QR code.

Image deformation techniques have been widely used in the entertainment industry, animation and medical imaging. Motivated by the fact that the lengths of some feature curves should remain constant during the deformation, an image deformation method is proposed with length-preserved feature curves. Intrinsic representation of the curves is used to achieve length-preserving curve deformation, and then the deformation of the image is driven by the deformation of the feature curves.

According to the nature of saliency map generation with colour contrast, a non-attention region first initialisation (NARFI) k-means clustering for saliency detection is proposed. The NAR is obtained by multiwavelet reconstruction based on the cutoff low-frequency. The initial seeds of the k-means are chosen from the NAR. This way, the NAR is clustered in a fine manner, whereas the attention region is clustered in a coarse manner. As a result, the saliency values of the attention region with the NARFI k-means clustering are more conspicuous than those with the k-means++ clustering.

At microwave frequencies, absorption by molecular resonance is a major factor affecting radio propagation. Irregularities in the radio propagation pattern, expressed in a form of the received signal strength indicator's (RSSI) variations, can indicate the possible presence of a human within the radio network. The proposed human presence detection method is based on the information entropy calculated over a set of principal components extracted from a sequence of RSSI samples incrementally, without estimating the covariance matrix. By applying the entropy algorithm, the information on human presence is quantified from the sequence of principal components. It is shown that through-the-wall human activities, which introduce disturbances in the RSSI footprint of the monitoring room, do not affect the detection accuracy of the method. Experimental results obtained for the 2.4 GHz indoor radio network assess the feasibility of the proposed approach.

A method to determine the minimum gain of the system is presented which can be identified by pseudorandom binary sequence (PRBS)-based approaches when noise is present in the output signal. A formula for the minimum gain is derived in terms of required noise resilience from the discrete Fourier transforms of the PRBS signals and white noise. The formula allows an engineer to discount noise-affected low-gain portions of the transfer function. The calculated minimum gain is compared graphically with noisy Bode plots and good agreement is shown with the actual noise level, demonstrating the usefulness of the result.

An improved baseband digital predistortion (DPD) solution to reduce the nonlinearity of power amplifiers (PAs) with memory effects is proposed. It is based on the memory polynomial (MP) model and linearly interpolated look-up table (LUT). This predistorter reduces significantly the computational time needed in MP based DPD by using a LUT. A linear interpolation technique is introduced to enhance the indexing of the LUT, such that good performance with small table size and low computational complexity is achieved. The proposed solution is validated on a test bench by using a PA ZFL-2500 driven by a 16QAM modulated signal.

High sensitivity AlGaN/GaN lateral field-effect rectifiers (L-FERs) were realised that can be monolithically integrated with AlGaN/GaN high electron mobility transistors (HEMTs) circuitry. F⁻ ion plasma treatment was optimised to achieve strong nonlinearity of the L-FER at zero-bias. The L-FER treated by F⁻ ion plasma at 120 W for 200 s exhibits a record level zero-biased curvature coefficient, γ, of 69 V⁻¹ at room temperature. The direct-measured voltage sensitivity is as high as 6.75 mV/μW at 2 GHz. The high sensitivity and AlGaN/GaN HEMTs compatibility of the L-FER make them very promising for zero-bias square-law detector applications.

A novel method of miniaturising the size of a coaxial cavity resonator by utilising zeroth-order resonance (ZOR) characteristics is presented. This miniaturisation has been achieved through the introduction of air-core coils inside a conventional coaxial cavity resonator. This structure possesses ZOR characteristics, which results in the miniaturisation of the resonator while still exhibiting a reasonable high unloaded Q factor. A duplexer was designed and fabricated using the proposed resonator. The measurement results verified that a 30% reduction in size was achieved compared with a traditional duplexer.

A miniaturised rat-race coupler capable of suppressing the second and third harmonics is proposed. The coupler is designed with slow-wave synthesised transmission lines using two π-networks in cascade. Each π-network is formed by a parallel LC tank and two shunt capacitors, and is associated with a transmission zero for harmonic attenuation. The two transmission zeros embedded in the synthesised line can be controlled independently. The proposed rat-race coupler, centred at 2.4 GHz with an operational bandwidth of 18.8%, and at least 21-dB suppression at the second and third harmonics, features a compact footprint, which is only 12.9% the size of its conventional counterpart.

A novel compact microstrip lowpass filter is described that exhibits an quasi-elliptical function response with high out-of-band rejection for integration in ultra-wideband (UWB) chipless radio frequency identification (RFID) receivers. The filter's characteristics include a wide-stopband and low passband insertion loss with a passband ripple of 0.03 dB. Out-of-band attenuation of 30 dB is achieved up to about four times the cutoff frequency (4f _c), and 20 dB up to about seven times the cutoff frequency (7f _c). In addition, the proposed filter structure offers a high figure-of-merit of 8657 that exceeds the lowpass planar filters with an equal cutoff frequency reported in the literature to date.

A terahertz frequency selective surface (FSS) which operates inside the spatial filtering front-end of a high sensitivity receiver is proposed and experimentally verified. To achieve this terahertz FSS metamaterial which exhibits extremely low insertion loss and is insensitive to the variations of polarisations and incident angles, the FSS array is constructed by an electrically thick SiO₂ layer (quartz wafer) and an ultra-thin high conductivity aluminium layer with periodic aperture unit cells. The micro-machined metallic pattern is shaped by the combination of a sub-wavelength circular ring and a split-ring resonator. Experimental results and predicted spectral responses are shown to be in good agreement over the frequency range of 0.1–0.5 THz.

A method for enhancing the sensitivity of a fibre Bragg grating (FBG) sensor by all-optical signal processing is demonstrated. Degenerated four-wave mixing (FWM) has been used to perform frequency chirp magnification, which can be used for magnifying the wavelength drift of the FBG sensor. Ultra-highly sensitive measurement of the static strain has been experimentally demonstrated with strain sensitivity of 5.40 pm/μɛ. The sensitivity has been enhanced by a factor of five.

A Si wire arrayed waveguide grating designed to reduce the noise and transmission peak width and avoid systematic phase error generated at the curved waveguides is reported. A slab waveguide structure to remove stray light is used. 200 GHz spacing 16 channel devices were fabricated by ArF immersion. An improved crosstalk of −23 dB was obtained.

A high-efficiency fast-transient-response V ²-controlled boost converter with a small equivalent series resistance (ESR) capacitor is presented. The proposed boost converter can achieve a fast and stable load transient response without relying on the ESR of the output capacitor. The switching frequency of the proposed boost converter is 500 kHz for a nominal 2.5–4.2 V input and a 3.3–5 V output application. Experimental results show that the transient recovery time inheriting the advantage of V ² control is shorter than 6.4 µs with the load step from 250 to 20 mA and the highest efficiency can be up to 95.4%. The proposed boost converter has been fabricated with a 0.35 µm 2P4M CMOS process, the total chip area is about 1.5 × 1.5 mm², including the PADs.

The simplified circuit and modulation scheme is presented for a five-phase three-level voltage source inverter (FPTL-VSI). The simplified FPTL-VSI is a concatenation of a dual-buck stage and a two-level VSI (TL-VSI), and the space vector pulse width modulation (SVPWM) is used as the modulation scheme. Owing to the merit of the dual-buck stage which uses only four controllable switching devices, the TL-VSI can produce a three-level voltage waveform. By simplification of the space voltage vectors, a simplified SVPWM scheme can be obtained. Thus, the simplified FPTL-VSI reduces the hardware costs and improves the quality of the output voltage waveforms. Furthermore, the hardware implementation of the simplified SVPWM scheme can be easily realised. The experimental results prove the feasibility and validity of the proposed inverter and the modulation scheme.

Time-domain pulsed I–V measurements dedicated to characterising and modelling the time-dependent trapping phenomena of wide band-gap AlGaN/GaN high electron mobility transistors (HEMTs) are presented. The influence of temperature, electric field and their mutual interaction on trap activation and time constants are investigated and illustrated in the case of 2 × 75 × 0.15 µm² AlGaN/GaN HEMTs from the III–V Lab. These measurements show that the most influential parameter on trap activation is the electric field induced by the highest drain voltage V _DS-max reached during a pulse sequence, which leads to the drain current collapse. These characterisations are intended to develop and improve new nonlinear models of GaN HEMTs, taking into account the dynamics of traps for modulated signals.

A novel ultra-thin partial silicon-on-insulator (PSOI) LDMOS with n-type buried (n-buried) layer (NBL PSOI LDMOS) is proposed. The new PSOI LDMOS features an n-buried layer underneath the n-type drift (n-drift) region close to the source side, providing a large conductivity region for majority carriers to significantly improve the self-heating effect. A combination of uniform and linear variable doping (LVD) profile with highly initial concentration is utilised in the n-drift region, which alleviates the inherent tradeoff between specific on-resistance (R _on,sp) and breakdown voltage (BV), as well as achieving low power dissipation. With a 60-µm n-drift region length, the NBL PSOI LDMOS obtains a high BV of 940 V with a low maximum temperature (T _max) of 327 K at a power (P) of 1 mW/μm, which is reduced by around 56 K in comparison to the conventional SOI LDMOS using an LVD profile for the n-drift region (LVD SOI LDMOS). Moreover, R _on,sp of the NBL PSOI LDMOS is lower than that of the LVD SOI LDMOS for a wide range of BV from 400 to 900 V.

A radiofrequency (RF) fingerprinting strategy for the proper identification of wireless devices in mobile and wireless networks is proposed. The main objective is the extraction of the main parameters of the RF signal of the device by means of the preamble in the communication in the wireless network. A complete procedure for RF fingerprint data extraction is provided, and some experimentation is conducted for its validation.