A linear frequency modulated transducer was earlier proposed for use in surface acoustic wave (SAW) tags and sensors. This reported work demonstrates that the hyperbolically frequency modulated (HFM) transducer has significant advantages for such devices often operating in a wide range of temperatures. The HFM transducer is practically insensitive to wide temperature variations, which expand or compress signals in time. Owing to the exponential change of the varying period with the electrode number, the expansion of the length of all the periods is equivalent to just a shift in time and the compressed signal remains practically unchanged in shape, just slightly shifted. Such a shift has no importance for SAW sensors/tags, which usually operate on the difference of delays of the compressed peaks.

A new family of waveguide power splitters based on the use of connected bi-omega particles is presented. The new layout consists of a H-plane T-junction, whose ports are characterised by the same cross-section. The output ports are both closed with a metallic screen and energy is allowed to pass through by means of a vertical slit drilled into each metallic screen. To make the transmission efficient, two identical particles, printed on the two faces of a dielectric slab, are placed across the slits. Depending on the design, at the resonant frequencies of the two bi-omega particles, either balanced or unbalanced power splitting is obtained between the two output ports. A proper set of numerical simulations and measurements confirm the effectiveness of the proposed design, which can be successfully used for realising compact and highly selective waveguide power splitters.

A new high-gain cylindrical dielectric resonator antenna (DRA) with a large bandwidth is proposed. A cylindrical dielectric resonator, an intermediate layer and a metallic cylinder are used to obtain a large bandwidth and a high gain. The intermediate layer structure can suppress surface wave propagation, which results in a gain improvement, and the cavity formed by the metallic cylinder provides a further higher gain. The measured results demonstrate that the proposed DRA achieves a large bandwidth of 23% from 5.4 to 6.8 GHz with a voltage standing wave ratio of < 2 and a high gain of around 11 dBi.

Two examples of fabric based frequency selective surfaces (FSSs) are presented. The FSSs are produced by using screen printing and weaving. Both measured and simulated data are presented showing excellent agreement and performance for the FSSs when compared with the simulated data. The performance of these samples points towards a useful screening technique using fabric hangings and wall coverings in a range of applications where temporary electromagnetic wave ingress or egress needs to be controlled.

A subspace-based method is considered to deal with multiple near-field source localisation with a uniform circular array. The property of the centro-symmetric circular array configuration is employed and a matrix which is free of range is constructed. Then, the two-dimensional (2D)-MUSIC method is used to search the 2D angles after the range parameters are separated. After the 2D angles are estimated, a 1D-MUSIC method is used to estimate the corresponding ranges when the estimated 2D angles are substituted into the original form of the received data. It is shown that the proposed method can deal with the case of multiple near-field sources, but with lower computational complexity than that of the available 3D MUSIC method; moreover, the 3D parameters are automatically paired.

A near-field switched loop antenna array with a circular defected ground structure for precise positioning systems is presented. By switching the loop array, the positions of the densely placed small-sized radio frequency identification tags can be identified. The characteristics of the proposed array are verified by the comparisons to the simulated results and experimental measurements.

A reconfigurable antenna with polarisation diversity is proposed. The proposed antenna consists of a square radiating patch and a switchable Y-shaped feed structure. The radiating patch is fed by the switchable Y-shaped feed line to dynamically change the polarisation state. The novel structure described here enables matched input optimisation for both polarisations. Low cross-polarisation levels in the linear state and good axial ratios in the circular state are observed.

A novel combination of a two-port slant ± 45° dual-polarised microstrip patch antenna with a polycrystalline silicon (poly-Si) solar cell operating as an RF stacked parasitic patch element is presented and discussed for 5.8 GHz band polarisation diverse multifunctional WiMAX systems. To minimise the effect of the solar operation on the RF antenna performance, an effective DC/RF isolation circuit consisting of quarter-wavelength microstrip transmission lines is demonstrated. The proposed solar patch provides identical E-plane and H-plane far-field radiation patterns for +45° and −45° polarisations with a gain of 7.8 dBi and operates at the frequency band of 5.66–5.91 GHz.

A flexible meandered loop antenna is proposed, designed and realised for small cylindrical implantable devices. This antenna covers the Medical Device Radiocommunications Service (MedRadio) and the 433–434 MHz industrial, scientific and medical (ISM) bands, and is probably the smallest reported flexible implantable antenna that covers these bands. This antenna has shown a robust performance in the presence of implant internal components and has a realised gain of − 28.4 dBi inside a model of a human upper arm.

The digital read-out scheme in an ion-sensitive field effect transistor (ISFET) sensing array is more advantageous when compared with the analogue counterpart because of its lower power consumption, smaller area and less susceptibility to environmental noise and parasitics. An improved read-out scheme is proposed in which each ISFET is stacked with a CMOS inverter to form a pH-to-time converter. The pH level of the solution regulates the strength of the ISFET, which in turn modulates the delay of the stacked inverter and hence the pulse width of the output signal. Simulation results using the 0.18 µm/2.5 V CMOS process show that the modulated pulse width changes linearly over a wide range of pH. The design achieves five orders of magnitude smaller leakage and 40% lower dynamic power consumption, while it requires only 50% of silicon area when compared with the conventional design. It is therefore more suitable for large ISFET arrays implemented in nano-scale CMOS technologies.

A novel method for restoring photoplethysmography (PPG) signals using electrocardiography correspondences and support vector regression (SVR) is proposed. To enhance the PPG quality, five correspondences between the period signals of PPG and ECG were automatically detected, and their position and amplitude are independently trained for defining the estimator. The result shows that the quality of the PPG signal restored using the proposed method is greatly enhanced and becomes similar to that of the ECG signal.

A novel architecture of an asynchronous sigma–delta modulator with noise shaping is introduced. Compared with the conventional solution, the proposed modulator achieves first-order noise shaping without an additional loop filter. With the same sampling frequency, the improvement of the SNDR of the asynchronous sigma–delta modulator can be over 20 dB.

A highly digital technique is proposed for excess loop delay (ELD) compensation in multi-bit continuous-time ΔΣ modulators. A digitally controlled reference switching matrix is used to replace the power-hungry signal adder and extra DAC driving the quantiser, which are commonly used in ELD compensation. With the proposed technique, the feedback DAC is embedded in the quantiser and implemented by a few switches. Thus, it allows low-voltage and low-power operation.

Recently, it was found that flux φ and charge q are a pair of complementary basic physical attributes generating elementary circuit elements. With such a finding, two basic electrical circuit laws and constitutive relations of three fundamental circuit elements are rebuilt, upon which new dynamical equations describing dynamical circuits containing memristors are established. This leads to investigation of whether the memristor is a dynamic element.

A swap time-aware garbage collection (STGC) policy for the NAND flash-based swap system is proposed, which focuses on reducing the cleaning cost and improving the degree of wear-levelling. STGC calculates the cleaning index value of each block to select a victim block and the normalised value of the elapsed swap time of each valid page within the victim blocks to identify the hot valid page and cold valid page. Trace-driven simulations with a synthetic trace show that the STGC outperforms the existing garbage collection policies.

A novel technique that facilitates near-field (NF) scanning for electromagnetic compatibility-compliance testing is described. It performs measurements in a sequential way with the aim of discovering multiple, possibly disjoint regions where the amplitudes of an NF component belong to certain output ranges. The measured data samples are used to train a classification model where each NF range is represented by a given class (e.g. low/medium/high NF amplitudes). The outcome of the algorithm is a visual map that clearly characterises and pinpoints the exact location and boundaries of each class. Such maps are useful, for example, to detect hotspots or regions that are prone to electromagnetic compatibility issues. The technique has been validated on a measured microstrip bend discontinuity.

An open-stub compensation technique is proposed for passive equalisers used in high-speed digital communication systems. By simply lumping an open stub to the conventional RL-type passive equaliser, the frequency-dependent channel loss from the long transmission path can be compensated, the frequency response of the channel becomes more flat and the bandwidth is wider. On comparing the proposed passive equaliser with typical RL passive equalisers, the simulation results show that the proposed technique can achieve an improvement by 1 GHz in the flat region of S ₂₁ on a 60 cm-long differential pair. Thus, the proposed technique successfully demonstrates an improvement in electrical performance of about 57.4% in the time domain.

The absorbing performance of a nonlinear waveform-dependent metasurface with pulsed signals is demonstrated. The metasurface is a periodic structure containing diodes, capacitors and resistors. These enable us to first rectify a high-power incoming signal to a static field, then store the energy during the illumination and dissipate it before the next pulse comes in. The incident pulses contain a finite width of spectrum of around 4.2 GHz. By using the nonlinear metasurface, absorption that depends on both the power and the duty cycle of the incoming signals is measured. These measurements demonstrate the first waveform-dependent absorbing metasurface.

An energy-efficient discrete cosine transform (DCT) is proposed. It can be used in image compression in wireless visual sensor networks. It is a combination of the recently proposed block discrete cosine transform and a pruned approach. Thus, the computational complexity is reduced significantly. The experimental results show that the proposed transform leads to a significant reduction in computation time at the target sensor node. This will consequently result in a saving in energy consumption. The proposed DCT is very suitable for implementation in wireless sensor networks powered by batteries.

Two multiplierless algorithms are proposed for 4 × 4 approximate-discrete cosine transform (DCT) for transform coding in digital video. Computational architectures for one-dimensional (1D)/2D realisations are implemented using Xilinx field programmable gate array devices. CMOS synthesis at the 45 nm node indicates real-time operation at 1 GHz yielding 4 × 4 block rates of 125 MHz at < 120 mW of dynamic power consumption.

A novel and efficient covariance-based method for person re-identification is proposed. The approach exploits three colourspaces and intensity gradients as covariance features and extracts multiple statistical feature vectors from the pyramid of region covariance matrices. The distance measure of the covariance pyramid is designed to be the weighted combination of four vectorised statistical features by cascading on the covariance pyramid. The method is compared with the state-of-the-art methods using a benchmark dataset and is demonstrated to outperform other state-of-the-art methods.

Lattice-matched InAlN/gallium nitride high electron-mobility transistors with a 6 nm-thick InAlN barrier layer were fabricated and characterised. By introducing a very thin InAlN Schottky layer, the short-channel effect could be minimised. The devices with a gate length of 75 nm exhibited output resistance as high as 56.9 Ω mm together with a drain-induced barrier lowering as low as 63 mV/V. The devices also demonstrated excellent high-frequency characteristics such as a unity current gain cutoff frequency (f_T ) of 170 GHz and a maximum oscillation frequency (f _max) of 210 GHz.

A wideband double-balanced mixer using 90 nm CMOS technology is demonstrated. The DC current paths of the transconductance stage and the switch stage are separated individually by an integrated transformer balun. The primary coil is absorbed in the DC path of transconductance and functions as an inductive load. Simultaneously, the secondary coils perform as a current source of the switch stage. The transconductance employs a common gate amplifier together with the gate-inductive gain-peaking technique for providing a wideband input matching and extending the operating bandwidth. The fabricated mixer operates from 15 to 40 GHz with around 5 dB conversion gain, matches to 50 Ω over the full band and only takes up 0.39 mm².

Drift-step-recovery diodes (DSRDs) are used as ∼ 1 ns high-voltage opening switches by pumping them slowly in the forward direction and then pulsing them quickly in the reverse direction. The fast opening occurs when the reverse current discharges the carriers that were stored in the DSRD junction during the forward cycle. Typical forward and reverse timescales are tens and a few nanoseconds, respectively. Although a state-of-the-art metal-oxide semiconductor field effect transistor may pulse the DSRD at a rise-rate of about 20 A/ns, the DSRD itself can be used to pulse another DSRD at a rise-rate of about 60 A/ns. An enhanced performance by the proposed method, resulting in a high-voltage nanosecond pulse has been reported. This pulse was then further sharpened by driving a fast avalanche diode. A 6-kV, 130-ps rise-time, with a rise-rate exceeding 40 kV/ns circuit is presented.

A novel and compact microstrip ultra-wideband (UWB) bandpass filter (BPF) based on a triple-mode resonator is designed, fabricated and measured. The proposed resonator is composed of a coupled line with a half-pass characteristic, which is connected directly with each other at one end and is short-circuited to the grounding plane at the other end. A simple structure and design approach to the proposed filter is provided and demonstrated. To implement the design on a microstrip, the directly tapped-transition input/output ports are used to excite the coupled-line multi-mode resonator. The measured results are in good agreement with the electromagnetic simulated results. The proposed filter shows a 3 dB fractional bandwidth of 117.3% at the central frequency (f _c) with a return loss of better than 12 dB and an insertion loss of < 0.5 dB within the passband response.

A high-selective microstrip wideband bandpass filter (BPF) with an adjustable notched-band is proposed. First, a pair of open-ended stubs is tap-connected to a λ/2 uniform impedance resonator at two symmetrical locations with respect to its centre, and they are further electrically connected by a microstrip line section. Such an arrangement leads to the emergence of a notched-band owing to the out-of-phase cancellation of two dissimilar signal paths in a wide passband. Secondly, the electrical path connecting the two open-ended stubs is gradually tuned so as to adjust the notched-band to the desired frequency location. Finally, a prototype filter is designed and fabricated, and its measured results are provided to verify the predicted frequency response.

A tunable bandpass filter (BPF) with low-loss and enhanced selectivity based on controllable coupled negative resistance is presented. Owing to the use of the coupled negative resistance, the insertion loss and the selectivity are improved significantly from 800 to 1000 MHz. Furthermore, a novel method of introducing an open-circuited stub and a short-circuited stub to control the negative resistance effect is proposed, which makes the adjustment of the negative resistance much more flexible. The measured active BPF shows a bandwidth of 56 MHz, an insertion loss of − 0.44 dB, a noise figure of 2.6 dB, 1 dB compression point of 0 dBm and third-order intercept point of 8.8 dBm at 900 MHz.

The first 40 Gbit/s multipath data routing is demonstrated in a two-dimensional switch matrix composed of high-order resonant optical switches. Broadband connectivity with on-state bandwidths exceeding 100 GHz, chip losses below 9.0 dB and signal extinction above 20 dB is achieved through the four paths on the diagonal of the 4 × 4 matrix. 40 Gbit/s data routing is demonstrated through three of these paths with power penalties below 0.5 dB for each path.

A simple and novel spectroscopic photo I–V method of diagnosing the homogeneity of electrically-active defect distribution in the large area AlGaN/GaN HEMT (high electron mobility transistor) epi-structure grown on 6-inch silicon wafers is reported. It is of utmost importance to produce the HEMT epi-structure electrically homogeneous across the wafer if devices with uniform electrical characteristics are to be constructed. AlGaN/GaN HEMT epi structures were grown on a silicon substrate via metal–organic chemical vapour deposition. An array of circular semi-transparent Ni Schottky contacts was prepared on top of the diced AlGaN/GaN HEMT structure substrates, which were selected from different locations of the 6-inch wafer. The information of the electrical homogeneity across the wafer was elucidated from the spectral dependences of the I–V characteristics collected from different locations of the AlGaN/GaN HEMT wafer. It is successfully demonstrated that the proposed spectroscopic photo I–V measurement technique can be employed to diagnose electrical homogeneity of the electrically-active defect distribution in the AlGaN/GaN HEMT epi structure constructed on Si with minimum sample preparation steps.

The possibility of reducing the relative intensity noise (RIN) of a quantum cascade laser is demonstrated. By phase-dependent optical feedback from a short, piezo-controlled external cavity configuration, it is possible to achieve up to 9.5 dB of RIN reduction. These findings are consistently explained by a detuned loading approach based on modifying the alpha parameter.

Improvement of the removal efficiency of the upper (directed towards the n-side) heat flux from high-power laser diodes is proposed by providing the graphene-oxide (GO) deposit from the top of the p-down-mounted chips. Resulting decrease of the device's thermal resistance has been found in reduced red shift of the spectral characteristics connected with the drive current increase (compared with ‘no GO deposit’ device). Providing GO in the form of aqueous suspension and then drying at room temperature is relatively easy and repeatable. Experimental step-by-step processing and characterisations indicate dominant heat removal from the chip's side walls rather than from its n-contact surface.

Single-mode operation of fibre-pigtailed distributed feedback semiconductor lasers in the 2.05 μm range has been demonstrated. The lasers are packaged inside standard butterfly modules with output powers in excess of 10 mW at the end of polarisation maintaining optical fibre. The fibre-pigtailed lasers show excellent sidemode suppression ratios ( > 50 dB) and have a mode-hop free tunability larger than 1 nm. The output of the optical fibre has linear polarisation with better than 20 dB extinction over the operating current and temperatures.

Conventional fluorescence-detection systems typically involve a separate laser excitation source shining directly down towards the photodetectors, with the fluorescent sample in-between. A major issue with such systems is the strong excitation light signal that can saturate detection circuitry and/or make detection of the fluorescent signal difficult, even with optical filtering. A fluorescence-detection system is described consisting of light emitting diode dies bonded on top of a CMOS imaging chip. Such a system facilitates a high level of integration while potentially reducing the excitation light hitting the photodetectors. Model results are used to predict the shape of the imager response to excitation light. Experimental results confirm the validity of the model.

A novel signal processing-based approach for the real-time estimation of air-gap length in fibre optic extrinsic Fabry-Pérot pressure sensors is presented; the algorithm is based on eigendecomposition of the spectrum mutual correlation matrix. A pressure accuracy of 6.1 Pa (0.045 mmHg) is achieved in a pressure probe with 1.13 nm/kPa sensitivity, highlighting a 6.4 times improvement over standard quadrature point amplitude tracking by just acting on signal processing.

A constant frequency control of an LLC resonant converter is presented. A switched capacitor is introduced to change the resonant capacitance, which enables the control of the voltage gain at a fixed switching frequency. The operation principle of the proposed converter with a switched capacitor is discussed. The experimental results are provided to show the validity of the proposed method.

A frequency-agile load modulation scheme is proposed for reliable near-field magnetic communication with a magnetic resonance wireless power transfer system. A working prototype of the system to realise concurrent power and data transfer in the high-frequency band has been fabricated to verify that demodulation of the load modulation signal state detected on the transmitter side is stable over the distance ranges covering the near-field coupling region. This scheme can be a simple and efficient solution against potential issues such as non-communication and envelope inversion arising from strong magnetic resonance coupling.

Thermal imaging of solar cells is important for diagnosing non-uniform operation or point defects, which can reduce cell efficiency. However, imaging with infrared light is impractical for superstrate CdTe cells because the glass substrate blocks transmission of light. It is shown that thermoreflectance – a lock-in technique that detects changes in the reflectivity of visible light – can circumvent this problem and achieve thermal images with spatial resolution limited only by the imaging wavelength. The diagnostic is used to show that a particular defect is a resistive shunt.

Two advanced fabrication methods are introduced: the hydrogen peroxide (H₂O₂)-added chemical bath deposition technique and molybdenum (Mo) back-contact formation under 3 kW sputter power. The parameters of the short circuit current density (J _SC) and conversion efficiency (η) were improved over standard cells when the ZnS buffer layer was deposited in the H₂O₂-added chemical solution. Otherwise, the fill-factor and η were best at 3 kW Mo sputtering power conditions. Advanced fabrication methods are realised to improve cell performance without modifying the chemical composition of the Cu(In,Ga)Se₂ absorption layer.

The micro-Doppler effect in inverse synthetic aperture radar (ISAR) imaging for aircrafts may impair the imaging quality of the main body, which affects application in target discrimination. The current available micro-Doppler effect reducing techniques are mainly complex and do not meet the requirement of real-time process. In the reported work, the profile flicker of micro-Doppler echoes is investigated by theoretical analysis and simulation. With the utilisation of this feature and based on compressed sensing, a new method which can reduce the micro-Doppler effect by preserving the information of aircrafts simultaneously is proposed. This method avoids the transform of the data field and the calculation is much less than the existing techniques. The validity is demonstrated by real data processing results.

The design of radar coherent pulse trains encoded with interval staggering for ambiguous Doppler target detection in clutter is addressed. To alleviate the Doppler ambiguous problem, an interpulse phase coding optimisation model is formulated in terms of staggered interval sampling, while maximising the output signal-to-interference-plus-noise ratio for improved target detection. Based on a reformulation of this optimisation problem as a non-convex quadratically constrained quadratic program (QCQP), a semidefinite relaxation algorithm for this QCQP is presented. Numerical results demonstrate the effectiveness of this proposed approach.

Phase unwrapping techniques remove the modulus 2π ambiguities of the wrapped phase. A robust phase unwrapping algorithm is proposed which reduces the probability of generating spurious phase jumps. Simulation results confirm that the new algorithm has better performance than the existing method at a low signal-to-noise ratio.

A pipelined circuit to calculate linear regression is presented. The proposed circuit has the advantages that it can process a continuous flow of data, it does not need memory to store the input samples and supports variable length that can be reconfigured in run time. The circuit is efficient in area, as it consists of a small number of adders, multipliers and dividers. These features make it very suitable for real-time applications, as well as for calculating the linear regression of a large number of samples.

An i-vector representation based on bottleneck (BN) features is presented for language identification (LID). In the proposed system, the BN features are extracted from a deep neural network, which can effectively mine the contextual information embedded in speech frames. The i-vector representation of each utterance is then obtained by applying a total variability approach on the BN features. The resulting performance of LID has been significantly improved with the proposed BN feature based i-vector representation. Compared with the state-of-the-art techniques, the equal error rate is relatively reduced by about 40% on the National Institute of Standards and Technology (NIST) 2009 evaluation sets.

Physical cell identity (PCI) is used to identify each cell. However, with the introduction of femtocells, 8 bytes allocation for PCI cannot accommodate a large number of femtocells. The current solution in 3GPP Release 9 is to use cell global identity associated with PCI to resolve this problem. However, this solution may incur higher inbound handover failure. A dynamic PCI assignment based on cell activity level prediction is proposed. The system level simulation showed that the approach outperforms the existing approaches.

A sparse noise reduction method applied to Loran-C skywave delay estimation is presented, which is both simple and practicable. Signal sparsity is utilised to reduce the noise inside the Loran frequency band and significant improvement in the performance of estimation is achieved. The proposed method shows increased estimation accuracy with reduced computational complexity owing to no matrix inversion in the computation process.

Orthogonal frequency division multiple access (OFDMA) is a promising technology owing to its flexibility in subcarrier/subchannel allocation. Since most applications in OFDMA wireless networks generate bursty traffic, it is efficient and flexible to employ a random access mechanism in such networks. A single-radio multi-subchannel carrier sense multiple access with collision avoidance (CSMA/CA)-based random access protocol (called SRMC-CSMA/CA) is proposed for OFDMA wireless networks. By means of intermittent carrier sense and cumulative update of backoff counters, a terminal with a single radio can utilise multiple subchannels to transmit packets concurrently. Simulation results indicate that the SRMC-CSMA/CA protocol apparently improves the channel utilisation, especially when the traffic loads of the terminals are unbalanced.

A sensor system using ultra-wideband frequency technology and passive surface acoustic wave (SAW) sensors/tags has been demonstrated experimentally. The system operates with a frequency bandwidth of 500 MHz, which results in compressed RF pulses of about 2 ns duration, including just a few sinusoids with amplitude modulation. A correlation method is developed to measure the delay between two echoes with high resolution, avoiding the phase ambiguity problem. For temperature, deformation or other measurand a simple structure including only two reflectors is sufficient. This method is used in a system which simultaneously remotely measures a few temperature sensors with a resolution of 0.1°C. The operation of the system in a strongly reflecting environment (inside a metal box) is demonstrated.