The development of a dual polarised microstrip antenna for S- and C-band with improved characteristics is presented. It is dual linearly polarised at S-band and dual circularly polarised at C-band. Two orthogonal feeds have been conceived for proximity coupling at S-band. The other two composite feeds are aperture coupled at C-band. These generate left-hand and right-hand circular polarisations exclusively. The antenna reveals 3.5% or more impedance bandwidth at the S-band ports with an isolation better than 22 dB (the best isolation is ≈31.5 dB at the band-centre). It yields >21.2% impedance bandwidth and about 10.2% axial-ratio bandwidth at each of the C-band ports, both being improved compared to the earlier reports. Achieved isolation between the C-band ports is better than 18.5 dB (≈23.5 dB at the centre of the working band).

A highly compact and broadband antenna suitable for a prosthetic device to restore vision in the blind is presented. By embedding three straight sections of meandered microstrip patch line at the feed point, the proposed antenna gives three separate bands at 2.45, 4.2, and 5.8 GHz. The square patch loop is etched in the centre for miniaturisation. Three U-shaped patch stubs are utilised for frequency tuning. Three bands are merged together by adopting two square annular rings in the ground to obtain a wider bandwidth (2–8 GHz) incorporating unlicensed industrial, scientific and medical (ISM) and UWB bands.

This work explains the design and development of a low profile and compact single probe-fed metamaterial (MTM)-based circularly polarised (CP) antenna using the modified rectangular patch with two pairs of spiral strips and vias for 5.2/5.8 GHz WLAN applications. The intended CP antenna mainly consists of two pairs of radiators and each radiator is realised by using epsilon negative (ENG) transmission line (TL) to generate zeroth-order resonance. CP characteristics are achieved by placing two radiators for x-polarised wave are orthogonally aligned to the other two radiators for y-polarised wave. The antenna radiating element offers a compact size of 0.24λ₀ × 0.22λ₀ with a low profile of 0.04λ₀ at 5.5 GHz. To confirm the characteristics of the intended MTM CP antenna design, the designed antenna is fabricated and measured. Measurement results show that the antenna obtains −10 dB impedance bandwidth of 610 MHz (10.86%) which is 2.3 times larger than the standard microstrip patch antenna and axial ratio bandwidth of 140 MHz (2.54%) is obtained.

Switched-capacitor networks have been used as bandgap references (BGRs) to replace the area-consuming resistors in ultra-low power applications, incorporated with voltage doubler (2X) charge pumps to operate at low supply voltages. However, the open-loop charge pumps are sensitive to supply and load variations. Therefore, in this work, constant bias currents are added to the cross-coupled 2X charge pumps to reduce the supply sensitivity, and also to reduce the output ripples without using large capacitors. To further reduce the power consumption, a low-leakage sample and hold (S&H) circuit is used for an intermittent operation, the active/idle duty ratio of the BGR core is 1/15. Designed in standard 65 nm CMOS, the proposed BGR reaches a temperature coefficient of 64 ppm/°C consuming only 5.4 nW with a minimum supply of 0.45 V (simulated).

Continuous-rate all-digital reference-less clock and data recovery (CDR) circuit that utilises a sub-harmonic extraction scheme for wide-range frequency detection is presented. In the proposed CDR, the capture range of the frequency locked loop (FLL) is extended to the tuning range of digital controlled oscillator, thanks to the subharmonic extraction scheme. The frequency errors of FLL in lock state are within the tracking range of CDR loop. The prototype reference-less all-digital CDR, fabricated using a 40 nm CMOS technology, successfully detects 8–10 Gbit/s PRBS 2³¹ − 1 data and produces the recovered clock. The CDR consumes 29 mW from a supply voltage of 1 V for 10 Gbit/s input data. The measured RMS jitter of the recovered clock is 2.24 ps.

An automated formal verification approach for ensuring input completeness of NULL Convention Logic (NCL) circuits is proposed. NCL circuits have the benefit that they can operate in extreme environments where traditional synchronous circuits fail due to significant fluctuations in circuit timing. Input completeness is a critical property to ensure correct functioning of NCL circuits in extreme environments and therefore is required to be verified. Note that an NCL circuit can be functionally correct and still not be input complete, which could cause the circuit to operate correctly under normal conditions, but malfunction only when the circuit timing is substantially changed (e.g. operating in a very hot or cold environment such as outer space).

A mathematical model of a tumour growth cancer cell is investigated. A simple analogue circuit for mimicking the tumour growth cancer cell model is constructed for the first time. The stability of the proposed system is studied. The familiar chaos control technique, i.e. the non-feedback loop is performed to control the chaotic behaviour. The possible application of this study is discussed.

A small area shift register for narrow bezel display applications is demonstrated. The proposed circuit achieves the substantial reduction on the number of thin film transistors (TFTs) by sharing pull-up as well as pull-down nodes over multiple output pulses. In addition, there is no bootstrapping degradation caused by additional capacitors of pull-up TFTs due to separating TFTs between pull-up nodes. For four outputs per stage, the number of TFTs is reduced to 53%. The simulation results with low-temperature poly silicon TFTs ensure that all pull-up nodes of the proposed shift register with separating TFTs are boosted by larger than 12 V without any degradation.

This Letter introduces a new circuit topology of the simple electronically controllable voltage-mode bilinear two port (BTP) employing two operational transconductance amplifiers. The proposed two-port circuit is completely resistorless and contains only a grounded capacitor. It has a high-impedance input (a simple cascading of sections is possible) and allows an independent electronic adjusting of zero and pole frequency of the transfer function. The operation of the BTP circuit is demonstrated for two different locations of the pole and zero frequency coordinates. Moreover, a direct application of four BTPs in fractional-order integrators (order −0.11 and −0.33) with phase shifts −10° and −30° is shown. Design specification is verified by experimental measurements.

The artificial potential field algorithm for path control of unmanned ground vehicles formation in highway is proposed. First, the basic unit is derived from decomposition-iteration principle to improve formation stability and flexibility. Then, the virtual-leader potential field (VLPF) and vehicle-vehicle potential field are combined to generate the formation path in straight highway. Especially the VLPF function boundary is set as ellipse to adapt the highway environment. Finally, the formation in curved highway can be achieved through changing the ellipse's orientation and dimension. Simulation results demonstrate the effectiveness of the proposed algorithm.

Magnetic noise originating from the ionosphere may be estimated by using the total electron concentration data retrieved by the global navigation satellite systems. This provides a useful scheme for mitigation of magnetic noise coming from the fluctuating ionosphere, which is a major interference source in the magnetic anomaly detection applications.

Conventional zero-shot learning (ZSL) methods usually use class-level attribute, which corresponds to a batch of images of same category. This setting is not reasonable since the images even though belong to same category still have variances in their attribute items. To alleviate this phenomenon, the authors propose a novel method namely attribute relaxation (AR) to extend attributes from class level to instance level by adding a small variance matrix, which is more reasonable than traditional ZSL methods such as Semantic AutoEncoder that projects features from multi to one. Extensive experiments on four popular datasets show that AR can significantly improve the method using only class-level attributes, and verifies that AR can make the projected features in attribute space more discriminative.

This Letter develops a line-laser-based real-time visual measurement method to inspect pavement rut depths. The detection vehicle equipped with line-laser projector continuously sends, in its normal driving state, laser light onto the pavement, which is captured by a high-resolution camera on the vehicle, thus a sequence of pavement rut laser line images are created. The pavement rut laser line image is transformed to the convexity index image, in which the peak points of the pavement rut laser line are rapidly sought to obtain the rut laser curve with a single-pixel width. The relationship between pavement rut depth and the visual deformation value of measurement points on the rut laser curve is determined by the calibration blocks, and then the pavement three-dimensional (3D) rut depths can be automatically measured, 3D rut depth curves and 3D rut surface can be reconstructed. Some experimental results verify the effectiveness of the proposed method.

The free-riding and large difference between upload and download amount of resources is a fundamental problem in a peer-to-peer network. An incentive mechanism, which can be implemented in a distributed fashion, can solve this problem. Global contribution (GC) approach is one such mechanism, but its speed of convergence is slow. This letter proposes a new index named biased contribution index (BCI). It is proved that BCI always converges at a certain value. Simulation results show that it converges faster than the GC.

The user selection method based on chordal distance is widely used in multiuser multiple-input multiple-output (MU-MIMO) systems with block diagonalisation. It has throughput close to the capacity based method with low complexity. In this Letter, a user selection method with higher throughput and much lower complexity is proposed. To investigate the performance in a more practical channel scenario, 3D MIMO channels are used in simulations. Compared with existing user selection algorithms, the proposed method achieves good throughput performance and the lowest complexity order.

A secret key can be generated between two legitimate parties from observations of the common wireless channel. However, key disagreement usually occurs due to the estimation errors. Considering a guard band, the key disagreement rate (KDR) is firstly derived in a closed-form expression. Although increasing the size of the guard band can decrease KDR; the length of the raw key bit sequence is also shortened. To evaluate such an effect, the raw key generation rate is defined and derived. The simulation results verify the theoretical analysis and are useful for the determination of pilot power and guard band size.

A 340 GHz 4 × 4 detector array based on 0.18 µm gallium arsenide technology is presented. Each pixel is composed of a planar log-periodic antenna, a Schottky barrier diode, and a pair of bonding pads. The detector array is mounted on a silicon lens and assembled onto a PCB using the authors’ novel flip-chip structure. Off-axis performances of the lens-coupled detector array are demonstrated experimentally, which agree well with the simulated ones. The detector array shows a peak voltage responsivity of 1347 V/W and a noise equivalent power of 2.1 pW/Hz^1/2 at 340 GHz, exhibiting much better performance than the CMOS counterpart.

A broadband 3-dB power splitter based on ferrite cores is proposed for RF communication systems. This power splitter is developed from the transformer-type Wilkinson power splitter, where lumped components are detailed designed to achieve excellent input and output impedance match in an ultra-wide bandwidth (300 kHz–400 MHz). The power splitter demonstrates better RF characteristics in comparison with former designs. The voltage standing wave ratios of the input and output ports are <1.2, the isolation between two output ports is >20 dB, the amplitude imbalance is <0.04 dB, and the insertion loss is <0.5 dB. Such power splitters have small dimensions and very low costs.

A 28 GHz 5-bit switch-type phase shifter (STPS) using 0.18 μm CMOS technology is presented for 5G mobile communication applications. Utilising a process variation insensitive transmission line-based bridged T-type topology for 11.25°, 22.5°, 45°, 90° 1-bit phase shifters and high-pass/low-pass network switched-path-type topology for 180° 1-bit phase shifter, the 5-bit STPS achieves ultra-low root mean square (RMS) phase error of 1.7° and low RMS amplitude error of 0.59 dB at 28 GHz with intrinsic digital phase-control mechanism. The measured insertion loss of all 32 states is 15 ± 1.5 dB at 28 GHz with zero dc consumption.

A single-mode 1060 nm vertical-cavity surface-emitting laser (VCSEL) and single-mode fibre (SMF) based link with two-tap active pre-emphasis enabling error-free transmission of 50 Gbit/s NRZ-OOK over 1000 m is presented.

A five-level LLC resonant converter is proposed, which is composed of two three-level (TL) bridge arms, main/auxiliary transformers, resonant elements and a set of rectifier bridge. The DC side and AC side of the two TL bridge arms are parallel connected and serial connected, respectively. With different modulation methods, the converter has three different operation modes with low, medium and high voltage gains. Compared with the traditional TL LLC resonant converter, it has lower switching frequency and higher efficiency at low gain mode. Using the same resonant parameters, it has a higher voltage gain and is more suitable for wide output range. A prototype with 90–220 V output was built. The measured efficiency of the new topology was 1.9% higher than that of the traditional topology at low voltage gain mode, and three gain modes were validated.

The problem of multichannel synthetic aperture radar image recovery in the presence of sidelobe narrow-band interference (NBI) is addressed. The NBI steering matrix is accurately estimated with the interference signal received in the passive mode. A beamforming technique is then utilised to minimise the energy loss of the target echoes and suppress NBI to a satisfactory level. By establishing a linear model for the signal with interference removed, the radar image is reconstructed via sparse recovery. The experimental results demonstrate that the proposed method can both suppress NBI effectively and generate high resolution images.

A novel scheme with the method of subcarrier modulation is proposed. Comparing with the other existing self-cancellation methods, the proposed scheme not only has the advantages of good inter-carrier interference (ICI) suppression capability and high bandwidth efficiency, but also shows a better bit error rate performance over multipath fading channels.