The feasibility of digital tuning of a programmable metamaterial-based antenna is explored. A leaky-wave antenna is implemented by cascading metallic J-shaped units in a composite right–left handed formation to create tunable metamaterial properties in terms of the radiation direction. The gaps in the J-shaped units are programmed in either open or close to alter the radiation patterns. Owing to the electromagnetic homogeneity where the unit length is smaller than one-fifth of the wavelength, changing the gap states has an equivalent effect to gradually tune the dispersion diagram. Eight J-shaped patches are integrated in the leaky-wave antenna, and nine different gap states are designed. The dispersion diagram is found to vary the transition frequency from 9.4 to 8.7 GHz with steps of 0.1 GHz, thereby frequency-dependent scanning radiation patterns are realised. Moreover, thanks to the gradually varying dispersion diagram, a state-dependent which also means programmable scanning radiation patterns are acquired in the band from 9 to 9.3 GHz. Experiment results agree well with the finite-element-method numerical analysis.

A compact coplanar waveguide (CPW)-fed triple-band monopole antenna with high-band isolation for various wireless applications is presented. The proposed antenna is composed of a CPW-fed structure and a hexagon-shaped patch on which a hexagon-shaped slot is etched for covering the wireless local area network (WLAN) band (2.4/5.2/5.8 GHz), and occupies a very compact size of 26 × 25 mm². By introducing a half-hexagon-shaped stub, additional bands can be realised covering the worldwide interoperability for microwave access (WIMAX) band (2.5/3.5/5.5 GHz). The proposed antenna has been fabricated and measured; results show a higher isolation between adjacent bands, good radiation pattern characteristics and stable gains in the operating bands. The simple feeding structure, compact size and uniplanar design make it easy to be integrated within portable devices for wireless communication.

A single-arm helical antenna with the width of the arm varying periodically for a circularly polarised (CP) tilted beam is introduced. The proposed antenna radiates a CP tilted beam, when the following two conditions are met: (i) the circumference of the proposed antenna is >4λ ₀/3 (λ ₀ is the wavelength in free space at the centre frequency) and (ii) the width of the arm varies periodically. The designed procedure reveals that the maximum radiation direction of the tilted beam is not sensitive to a change in frequency. The simulated and measured results show that the designed antenna radiates a CP tilted beam with θ _max = 32°. The measured 10 dB return loss bandwidth is 12.8% from 2.33 to 2.65 GHz, and the experimental results of the radiation pattern, gain and axial ratio are also presented.

An implantable glucose monitoring device based on a CMOS line sensor combined with a glucose-responsive fluorescent hydrogel is developed. An implantable glucose monitoring device with sufficient sensitivity and response time is realised, which is an improvement over the authors' previous proof-of-concept device. A reduction of the sensor size made the device diameter injectable with a commercially available 16-gauge syringe needle. The sensor's performance was confirmed through both in vitro and in vivo experiments.

A frequency-domain algorithm for the early detection of lung cancer is presented. The algorithm predicts the distribution of scattered fields inside the imaged domain (torso) using the measured fields around that domain. That prediction is based on using the first-order Bessel function of the first kind to relate the fields outside the imaged domain to the fields inside that domain. The predicted field distribution shows the relative differences between the dielectric properties of tissues within the torso and thus enables detecting lung cancer, which has a significantly larger dielectric constant that the lung's healthy tissues. To validate the proposed algorithm, an integrated imaging system, which includes a three-dimensional slot-rotated antenna that circularly scans an artificial torso phantom using the band 1.5–3 GHz, a wideband microwave transceiver and a laptop for control, processing and image generation, is built. The obtained experimental results confirm the reliability of the proposed method in lung cancer detection.

An efficient replica bitline (RBL) technique for reducing the variation of sense amplifier enable (SAE) timing is proposed. Both RBLs and four-fold replica cells compared with the conventional RBL technique are utilised to favour the desired operations. Simulation results show that the standard deviation of SAE can be suppressed by 44.25% and the cycle time is also reduced by ∼30% at a 0.8 V supply voltage in TSMC 65 nm technology. Additionally, the area of the proposed scheme is nearly the same as that of the conventional RBL scheme.

A class EF2 power oscillator designed in standard 130 nm CMOS at 2.5 GHz frequency is presented. The oscillator relies on a direct path based on a power amplifier and a feedback path based on passive elements and an MOS varactor. Class EF2 is used to reduce voltage stress across the switch, enabling a higher output power for modern transistors with low breakdown voltage. The measurement on a class EF2 power oscillator at radio frequency (RF) is presented for the first time. The circuit achieves 17.65 dBm output power from a 2.5 V supply voltage with 27.1% DC-RF efficiency and presents a 150 MHz tuning range. The measured phase noise is −101.6 dBc/Hz at 1 MHz offset. The circuit was implemented in standard 130 nm CMOS technology and consumed a total area of 1.95 mm². To the authors' knowledge this class EF2 power circuit has never been presented either at RF frequencies or in an integrated technology.

An analogue equaliser with a novel digitally tuned variable degeneration resistance is realised in a 65 nm CMOS technology. Implemented with three parallel resistance branches and one serial resistance branch to well fit the optimal conductance curve, the proposed variable degeneration resistance is exploited to achieve a wide boost range while significantly improving the linearity of the tuned boost factors compared with the equaliser with a traditional degeneration resistance structure.

A coaxial cable drive mechanism is proposed to implement a high reduction gear ratio. The cable drive mechanism has great advantages such as lightweight, zero backlashes and high degrees of freedom in terms of mechanical design. However, it is restrictively utilised in robotic applications because it is difficult to implement a high reduction gear ratio of more than 10:1 due to the complexity of adjusting tension components, which causes the mechanical structure to be rather bulky and heavy to be mounted. To overcome this, in the present author's previous work a multi-level reduction mechanism, by adopting a seamless winding method which links the previous output axis to the next input axis, was developed. The performance of the previous work, by adopting a coaxial mechanism to drive the multi-level reduction system is improved. Significantly, the proposed mechanism reduces the overall size due to efficient integration of mechanical parts. On the basis of the proposed mechanism, a three-dimensional computer-aided design is provided and the prototype is manufactured, where the usability of the prototype is verified by experimental results.

Recently, the low-rank representation (LRR) has been widely used in computer vision and pattern recognition with great success owing to its effectiveness and robustness for data clustering. However, the traditional LRR mainly focuses on the data from Euclidean space and is not directly applicable to manifold-valued data. A way to extend the LRR model from Euclidean space to the Stiefel manifold, by incorporating the intrinsic geometry of the manifold, is proposed. Under LRR, an appropriate affinity matrix for data on the Stiefel manifold can be learned; subsequently data clustering can be efficiently performed on the manifold. Experiments on several directional datasets demonstrate its superior performance on clustering compared with the state-of-the-art approaches.

According to a new market research report, electronic access control system is expected to be worth $16.3 billion by 2017. Vision-based biometric authentication systems have received much attention with increasing demands for long distance surveillance applications and access control to the security area. Such visual application is mainly focused on face recognition. Nevertheless, small size or poor quality images with varying poses, illumination, expressions, glasses or hats and so on can perturb the recognition. To deal with such problems, the fusing of both face recognition and gait analysis is proposed. Gait has recently become one of the most emerging biometrics for non-intrusive person identification. A system is proposed in which information provided by face recognition and gait analysis is fused to identify people. Such a system has direct industrial applications since it is a non-intrusive security system that is reliable, easily deployable and just needs an inexpensive camera.

An efficient three-dimensional (3D) object registration method considering the point cloud distribution for a mobile robot platform is suggested. This technique adjusts overlapping areas between two consecutive scan data according to the property of the observed object's surface. The proposed technique can reduce computational load and increase the accuracy of 3D object registration with less total number of scan data. The improved performance is verified by comparing the estimated pose with the ground truth pose for observation robots.

Contrast enhancement is an important step in many image processing and analysis applications. Although the idea of enhancing image contrast is simple, most published mechanisms of letting computers do so automatically tend to be complex or not so intuitive, and the performance is generally not consistent across different image shooting conditions. A new naturalistic approach to this problem that directly mimics what a human artist would do for contrast enhancement is proposed. The goal is to make every non-noise detail easily perceivable by a normal human. Specifically, the gradient/contrast and local background are computed for each pixel and then linearly mapped to an objective point in the gradient–background plane. Then, a cross-bilateral filter is used to smooth the two linear map parameter images. A moving local window is used to obtain the local noise level for each pixel, which is used to determine whether the pixel should be enhanced. The proposed approach is easy to implement and leads to superior results compared with typical state-of-the-art methods, as confirmed by experimental results.

The three-dimensional (3D) extension of the high-efficiency video coding (3D-HEVC) standard adopts new depth-modelling modes (DMMs) to provide an alternative prediction scheme for depth-map intra-coding. In 3D-HEVC, although edges in depth maps can be accurately estimated by utilising the DMMs, testing the DMMs in the mode decision introduces a huge computational load to the encoder. A fast mode decision algorithm is proposed that can significantly reduce the computational overhead incurred by the DMMs. Experimental results indicate that the proposed algorithm reduces the encoding complexity by 33.89% on average without any penalty in coding efficiency.

Owing to the validity of measuring the visual visibility for image distortion, just noticeable distortion (JND) has been widely used in the quantisation-based watermarking framework. However, the existing JND model treats every region in the image with an equal attention level. Visual saliency, which reflects the visual attention, is proposed to improve the perceptual JND model. Based on the improved model, a logarithmic spread transform dither modulation (STDM) watermarking scheme is proposed. Simulations show that the proposed watermarking scheme with the improved JND model has superior robustness compared with existing STDM schemes.

A novel method is proposed for pose-invariant gender classification based on three-dimensional (3D) face reconstruction from only 2D frontal images. A 3D face model is reconstructed from only a single 2D frontal image. Then, for each two-class of gender in the database, a feature library matrix (FLM) is created from yaw face poses by rotating the 3D reconstructed models and extracting features in the rotated face. Each FLM is subsequently rendered based on the yaw angles of face poses. Then, an array of the FLM is selected based on the estimated yaw angles for each class of gender. Finally, the selected arrays from FLMs are compared with target image features by support vector machine classification. Promising results are acquired to handle pose in gender classification on the available compared with the state-of-the-art methods.

 A new architecture for a polar decoder using a reduced complexity successive-cancellation (SC) decoding algorithm is presented. This novel fully unrolled, deeply pipelined architecture is capable of achieving a coded throughput of over 237 Gbit/s for a (1024, 512) polar code implemented using a field-programmable gate array. This decoder is two orders of magnitude faster than state-of-the-art polar decoders.

A frequency-weighted balancing related model order reduction technique is proposed that provides stable reduced-order models for the case when both input and output weightings are included. The reduced-order models are realisation independent. Comparison with other frequency-weighted balancing related model order reduction techniques reveals the superiority of the proposed scheme.

The impact of different kinds of interleavers on the performance of block Markov superposition transmission (BMST) systems is investigated. Simulation results show that the performance of a BMST system is insensitive to the types of interleavers, provided that multiple interleavers (if required) differ from each other. As a byproduct, a new interleaver called the binary linear interleaver is proposed, which can be easily implemented in hardware.

An inductorless vector modulator-based radio frequency (RF) weighting circuit working from 1 to 10 GHz is presented. Amplitude and phase interdependencies were strongly reduced employing optimised attenuators. The vector modulator features a 360° phase and more than 40 dB amplitude control range with a root mean square phase and amplitude error of only 6° and 0.16 dB, respectively. These features make it a suitable candidate for communication and localisation diversity frontends in the 2.4 and the 5.9 GHz industrial, scientific and medical (ISM) bands. The circuit was fabricated in a 180 nm BiCMOS technology and contains an inter-integrated circuit (I²C) control interface and digital-to-analogue converters. It consumes 87 mW and requires an area of only ≃0.2 mm².

A novel topology of a miniaturised dual-band branch-line coupler with a 3 dB power split is presented. The compact design is achieved by folding the transmission line sections into the interior of a circuit. The initial dimensions are obtained at the equivalent circuit level. Design closure is conducted at the electromagnetic-simulation model level. The coupler operates on frequencies of 1 and 2 GHz, and features miniaturisation of 75% compared with the conventional structure. Experimental validation of the fabricated prototype is provided.

A wideband phase inverter (WPI) integerated in a Doherty power amplifier (DPA) is proposed for linearity improvement. This WPI is integrated at the output of the carrier amplifier in the DPA while keeping the same size as the conventional DPA. The input signal is excited at both the carrier and the peaking amplifiers and then the amplified signals are combined by the push–pull toplogy. This results in the cancellation of the even-order harmonic terms as well as the cancellation of IMD3 by proper bias selection. An experiment shows the peak power-added efficiency (PAE) is around 50% at 21 dBm output power with linearity improvement as the adjacent channel leakage ratio (ACLR) under a wideband code division multiple access signal is around −40 dBc. This improves around 7 dB in ACLR under the same output power and PAE and better error vector magnitude (EVM) compared with the single-stage class AB amplifier.

In this reported work, a modified decoding and detection scheme with dispersion compensation is investigated for spectral-amplitude-coding optical code division multiple access (SAC-OCDMA) systems. For simultaneously decoding and compensating the chromatic dispersion of signals in the 40 km-long standard single-mode fibre transmission, based on reconstruction equivalent-chirp method the array configuration of fibre Bragg grating (FBG) segments can be controlled with a quadratic dispersion of −680 ps/nm for an extra time delay. The traditional triple-branch signal detection (TBSD) scheme is then modified and the principle of multiuser interference elimination is theoretically analysed. The simulated results show that, without adding a dispersion compensation device, the modified TBSD scheme adopting the FBG decoder can realise the decoding and detection function in SAC-OCDMA.

A new multiple multicore fibre (MCF) connector is presented that employs the compressive deformation of the fibre in a multiple ferrule. This connector enables a physical-contact connection for all the fibre cores including the outer ones with a small compression force even if the connector end has manufacturing errors. The connector has been designed and fabricated for a 7-core single-mode MCF connector taking account of the relationship between the connector end structure and compression force, and it achieves physical contact connection with a connection loss below 0.7 dB and a return loss exceeding 54 dB.

A vertically stacked quad junction photodetector (PD) with an enhanced spectral responsivity for blue light is presented. This PD is implemented in a 0.35 µm high voltage complementary metal-oxide semiconductor (CMOS) technology. In addition, an optical window, an antireflection coating and a modified p ⁺ doping are used to attain the enhancement for blue light. The proposed PD is compared against other state of the art implementations and characterised for its spectral response.

The performance of a printable, ultraviolet (UV) photoconducting biopolymer is investigated for UV photodetectors of varying layer thicknesses. The biopolymer is formed from deoxyribonucleic acid (DNA) with the addition of the Clevios P formulation of poly(3,4-ethylenedioxythiophene)-poly(styrenesulphonate) (PEDOT:PSS) and hexadecyltrimethyl-ammonium chloride (CTMA); it is then combined with phenyl-C61-butyric acid methyl (PCBM) to make a printable, UV photoconducting material. The highest measured responsivity of the photodetectors is 1.2 mA/W at 20 V bias using a 260 nm source.

A simple method for the determination of the capture velocity of charge carriers from a three-dimensional (3D) region (waveguide region) into a 2D region (quantum well) is proposed. The method is based on measurement of the threshold current density and internal differential quantum efficiency in a semiconductor laser structure. The method also allows determining the 2D carrier density in a quantum well, which is otherwise not easy to measure in a multilayer laser structure.

Micro-scale, reflective-type optical displacement sensors have been developed by monolithic integration of vertical cavity surface-emitting lasers (VCSELs) and resonant cavity-enhanced photodetectors (RCEPDs). This compact sensor (700 × 700 μm) consists of an oxide-confined VCSEL that is surrounded by an RCEPD and can measure the linear distance with a measurement range of ∼10 mm, which is comparable to that of conventional displacement sensors. Fabrication details and sensor performances are discussed.

Geosynchronous (GEO) synthetic aperture radar (SAR) has an orbit height of 36 000 km which results in long round-trip time. Thus, the ‘Stop–Go’ assumption and the conventional slant range model for low-Earth-orbit (LEO) SAR lose effectiveness in GEO SAR. A concise signal propagation delay equation of GEO SAR is presented and furthermore an equivalent slant range model is proposed. To verify the accuracy of the proposed model, the dichotomy method is applied to obtain the real slant range. In addition, the effects of the ‘Stop–Go’ assumption on GEO SAR are analysed by theoretical analysis. Promising results were obtained by several experiments.

The availability of an expediting back-projection (EBP) algorithm for circular synthetic aperture radar (CSAR) imaging is investigated. EBP works by dividing the full aperture into sub-apertures, and using a unified polar grid to generate sub-images. On the basis of the correspondence between the wavenumber domain and the phase history domain, EBP combines all sub-aperture spectrums into a full-aperture one and reconstructs the CSAR image followed by a two-dimensional (2D) fast Fourier transform. Without time-consuming 2D interpolation, EBP is capable of decreasing the loss of spectrum, while improving significantly the computational efficiency compared with direct BP.

Device classification is important in many applications such as industrial quality control, through-wall imaging and network security. A novel approach has been proposed to use a digital noise radar (DNR) to actively interrogate microwave devices and classify defective units using ‘radio frequency distinct native attribute (RF-DNA)’ fingerprinting and various classifier algorithms. RF-DNA has previously demonstrated ‘serial number’ discrimination of numerous passive radio frequency signals, achieving classification accuracies above 80% using multiple discriminant analysis/maximum likelihood (MDA/ML) and generalised relevance learning vector quantisation-improved (GRLVQI) classifiers. It has also demonstrated above 80% classification of limited active interrogation responses with a DNR signal using these classifiers. The performance capabilities of the two different classifiers, MDA/ML and GRLVQI, on RF-DNA fingerprints produced from the ultra-wideband noise radar correlation response is expanded.

Generalised frequency division multiplexing (GFDM) has gained significant importance as a contender for the fifth generation (5G) air interface. The peak-to-average power ratio (PAPR) of GFDM is high due to the use of a same pulse shaping filter per subcarrier and the addition of different subcarriers at the transmitter. Proposed is the use of discrete Fourier transform (DFT) spreading-based GFDM transmission to reduce the PAPR. It is found that DFT spreading helps to reduce the PAPR of GFDM significantly. It is also seen that the bit error rate performance, which is computed through simulation and using an analytical expression of the signal-to-interference-plus noise ratio, is not compromised by DFT spreading but rather is improved in the frequency selective fading channel.

A new least-square active element pattern expansion (LS-AEPE) method is presented. With the help of this method, the array's pattern including the element mutual coupling and platform effect can be efficiently calculated by the fast Fourier transform (FFT). These techniques are then combined with the genetic algorithm (GA) to optimise the excitation phases for a 64-element linear array mounted on a trapezoid metal plate.

An efficient multi-user multiple-input–multiple-output (MIMO) protocol is proposed and verified to be significantly advantageous over the conventional IEEE 802.11ac protocol for next-generation wireless local area network systems employing massive MIMO and orthogonal frequency division multiple access.

The automatic classification of the modulation format of a detected signal is the intermediate step between signal detection and demodulation. If neither the transmitted data nor other signal parameters such as the frequency offset, phase offset and timing information are known, then automatic modulation classification (AMC) is a challenging task in radio monitoring systems. The approach of clustering algorithms is a new trend in AMC for digital modulations. A novel algorithm called ‘highest constellation pattern matching’ is introduced to identify quadrature amplitude modulation and phase shift keying signals. The obtained simulation and measurement results outperform the existing algorithms for AMC based on clustering. Finally, it is shown that the proposed algorithm works in a real monitoring environment.