As is well known, the global navigation satellite system (GNSS) antenna always suffers from the size problem. The authors' propose that two shorting probes in parallel are loaded on a patch antenna and a dual-band compact GNSS antenna, which has a smaller electrical size of only 60 mm (0.23λ) × 60 mm ×18 mm (0.07λ) × (λ being the wavelength in free space). A dual-layer feed network is also designed to achieve circular polarised radiation. Simulated and measured results show that the proposed antenna has good and stable performance over both dual operating bands.

A low-profile antenna that includes an integrated artificial magnetic conducting (AMC) ground plane is presented and is fully fabricated using multi-material additive manufacturing. It is shown that combining standard fused deposition modelling with conductive micro-dispensing printing enable one to realise mechanically robust antenna systems that contain spatially variable dielectric properties and intricate three-dimensional conducting networks. To illustrate this capability, a low-profile antenna is designed, fabricated, and experimentally validated, consisting of a standard 2.6 GHz half wavelength printed dipole, balun, and AMC ground plane.

A compact substrate integrated waveguide (SIW) based wearable tri-band leather antenna, designed for optimal on-body performance in the 2.4 GHz (ISM), 3.51 GHz (WiMAX) and 4.69 GHz (Military application) bands is proposed. By using brass eyelets and a combination of conducting and non-conductive leather materials, a substrate integrated waveguide cavity with a ground plane is realised. The antenna is very compact, flexible and directs radiation away from the wearer. Additional miniaturisation is achieved by introducing slits. The free space performance in the bands is found satisfactory with measured impedance bandwidths of 11, 7 and 5%, maximal measured free-space gains of 1.1, 0.9 and 2.1 dBi, and efficiencies of 71.3%, 68.2% and 76.5% achieved for the proposed antenna.

A highly integrated bidirectional analogue front-end (AFE) circuit for interfacing capacitive micromachined ultrasound transducer in medical imaging systems is presented. The proposed AFE features a reconfigurable high-voltage (HV) pulser which generates over 15 V pulses at 2.6 MHz in the transmit mode and also operates as an HV isolation switch for the receive path circuits, thereby saving significant silicon area. The AFE also includes a low-power low-noise preamplifier which achieves 103 dBΩ transimpedance gain at 2.6 and 5 MHz bandwidth while operating at 1.65 V supply voltage. The AFE integrated circuit (IC) is implemented using 0.18 µm standard CMOS process and the total area of the single-channel core is 0.052 mm² which is more than 65% of size reduction in comparison to previous work.

Electroencephalography signals are used for computer-aided diagnosis of Alzheimer's disease. Therefore, extracting critical features that belong to Alzheimer's signals are useful and tedious for neural network classification due to the high-frequency non-stationary components. For this purpose, time–frequency analysis and the multi-resolution capability of wavelets represent an attractive choice. However, fluctuations of the transformed coefficients and the absence of phase information make the process less accurate in certain scenarios. Because of this, complex wavelet transform has been selected to handle Alzheimer's signals. Moreover, the importance of calculating an optimal threshold value has been highlighted, usually by means of Shannon entropy as a helpful threshold identifier of the complex wavelet transform used to produce significant results. The effectiveness of Tsallis entropy instead of Shannon entropy in handling Alzheimer's signals is evaluated, the former giving place to better features for neural network classification. As a result, accuracy has been improved from 90 to 95% using Tsallis entropy. Hence, this new proposal boosts the opportunity to reduce mortality rates by detecting the disease accurately.

The critical challenge in designing complex digital circuits automatically using an evolutionary algorithm is to overcome the scalability issue caused by combinatorial explosion. Because of this problem, the existing evolutionary design methods can deal with digital circuits composed of only a small number of bits. To resolve this problem, a novel evolutionary design method is proposed on the basis of hierarchical module structures and predominant component prevention. From extensive experiments, it is found that this method can automatically design digital circuits composed of a large number of bits within a polynomial time.

A fast algorithm for the computation of a cosine-like number-theoretic transform is presented. The method, which corresponds to a finite field extension of a method originally designed for computing real-valued discrete cosine transforms, is recursive and suitable for VLSI implementation. A general flow diagram for the proposed algorithm is given and shows that, in some specific cases, it can be evaluated using additions and bit-shift operations only.

A new image encoding scheme for simultaneous encryption and compression applications, which is based on random convolution and random subsampling, is proposed. In comparison with the existing joint optical encryption and compression schemes tailored for multiple images, the proposed scheme can process a single image and achieve a robust reconstruction. The scheme with the architecture of double random masks is somewhat similar to double random phase encoding. However, the latter has no compression capability and the proposed scheme remedies this defect.

In radar-based time-domain head imaging, a priori information on the effective dielectric constants from the antennas’ perspective with respect to different head pixels is needed to map the significant scatterrers inside the head. A microwave-based head imaging technique, which utilises a model for the effective dielectric constant as a function of the antenna's position and imaged point inside the head, is presented. To build that model, a realistic MRI-derived numerical head surrounded by an array of wideband antennas is implemented in a full-wave electromagnetic solver. The time difference of arrival between transmission coefficients in free space and inside the head is calculated at different head pixels and used to evaluate the effective dielectric constant at those pixels. The model is validated on two different head phantoms and used to accurately detect different brain strokes.

A CMOS-microelectromechanical system 0.18 µm process of the Taiwan Semiconductor Manufacturing Company is used to design an integrated chip of a capacitive three-axis accelerometer and a sensing circuit. The sensing architecture uses capacitive sensing principles, and the three axes are integrated on the same structure and share the same mass. The three-axis design method can not only reduce the chip area but also has good sensing ability. The readout circuit architecture combines two capacitor voltage converters and a subtraction amplifier, which can change the sensing signal into voltage amplitude values. As a result, the total chip area is 1.6 × 2.0 mm², and the sensing range is from 1 to 4 G. The simulated sensitivities of X, Y, and Z are 34.8, 20.0, and 13.7 mV/G, respectively.

The use of an artificial neural network (ANN) for extraction of a temperature profile from a local Brillouin gain spectrum in a Brillouin optical time-domain analysis fibre sensor system is proposed and demonstrated. An ANN is applied to process the Brillouin time-domain trace in order to extract the temperature information along the fibre after the data acquisition process. The results show that the ANN provides higher accuracy and larger tolerance to measurement error than Lorentzian curve fitting does, especially for a large frequency scanning step. Hence the measurement time can be greatly reduced by adopting a larger frequency scanning step without sacrificing accuracy.

A new technique for exciting the TE₁₁ mode in a number of concentric coaxial waveguides is described. The excitation network is log-periodic and is designed to integrate a frequency multiplexing structure together with suppression of higher order modes (TE₂₁ and TE₃₁). The excitation network enables the implementation of very wide band nested coaxial feeds for reflector antennas where multiple waveguides are fed from a common input port. An excitation network providing continuous coverage over a 1–8 GHz frequency range was built and tested in a three-band coaxial waveguide feed.

A novel MEMS-based tunable meander inductor is reported. The inductor is fabricated using a single metal layer in a low-cost commercially available Metal-MUMPs™ process. The mutual inductance of the meander inductor is tuned by changing the gaps between the meander turns using a chevron-type thermal actuator. An air trench underneath the inductor has been made to reduce the loss in the silicon substrate and improve quality factor (Q-factor). Tuning range of ∼50.8% and maximum Q-factor of 8.17 have been achieved at 5.79 GHz by stretching the meander inductor for 22 µm. The total footprint of the device is 0.58 mm².

A liquid crystal filled continuously tunable substrate integrated waveguide filter is proposed which has a high-quality (Q) factor. To measure the Q-factor, a single resonator measurement is performed by using magnetic biasing in a lab demonstrator for a proof of concept. The unloaded Q-factor is 102–105.6. The resonator is employed to create a three-pole Chebyshev filter with a centre frequency of 22 GHz and a bandwidth of 600 MHz. Measurement results show a 6 dB insertion loss with a return loss of about 10 dB. The filter is continuously tunable within a tuning range of 610 MHz.

A tri-band bandpass filter (BPF) with an ultra-wide stopband is proposed. Based on tri-mode unequal-length shunted-line stepped impedance resonators (SIRs), some extra resonator sections are added in unequal-length shunted-line SIRs to increase the degrees of freedom in extracting the coupling coefficients of the tri-band BPF. Unequal-length shunted-line SIRs can provide an efficient way to obtain a tri-band BPF with a small frequency ratio between the first two centre frequencies. Moreover, two transmission zeros can be achieved near every passband by employing cross-coupling, resulting in high passband selectivity, excellent band-to-band isolation and an ultra-wide stopband. As an example, a tri-band BPF is demonstrated. Simulated results of the filter agree well with the measured ones.

Although several companies are selling MMIC distributed amplifiers with indications of two or more decades of usable bandwidth, the data has been difficult to implement in a practical amplifier design. Suppliers generally make coupling C's and bias tee's an external part of the MMIC measurement system and their losses are often removed (de-embedded) from the published data. Application engineers for even the largest vendors often seem to be unaware of how to practically achieve the bandwidth properties inherent to their devices, which requires simultaneous optimisation of both the high- and the low-frequency response. In this reported work using the axiomatic design approach allows systematically eliminating bandwidth bottlenecks and obtaining greater than two decades of bandwidth in a high-power module. The provided example uses an internally matched GaN distributed MMIC, but the same techniques are useful in many other configurations.

The design, simulation and measurements of microstrip slow-wave lines implemented in a 0.25 µm SiGe bipolar CMOS (BiCMOS) process is addressed for Ku-band (10–15 GHz) applications. The simulation results and the measurements show better performances for the proposed microstrip slow-wave line compared with the classical microstrip transmission line. These lines present very low insertion losses and a high phase constant.

The architecture and performances of a multilevel driver for pulse amplitude modulation (PAM) formats, designed and fabricated in 0.7 µm InP double-heterojunction bipolar transistor technology, are reported. The driver part is based on a power-DAC architecture which is integrated with the multiplexing stage composed of three 2:1 selectors. Up to 100 GS/s operation was validated and PAM-2, -4, -8 signals with high amplitude were measured. In particular, PAM-4 at 84 GBd and PAM-8 at 64 GBd operation was demonstrated with, respectively, a 3.7 and 4 V_pp differential output signal. This compact driver circuit is characterised by the highest merit factor in terms of high amplitude and the transmission capacity for an electronically generated multilevel signal.

The performance assessment of a novel multilevel modulation format based on partial-response signalling called absolute added correlative coding (AACC) by numerical simulations is delineated, targeting short-range and interconnect high-speed optical networks. The spectral efficiency, chromatic dispersion tolerance and receiver sensitivity of 4-instensity level AACC are discussed and compared against other well-known existing modulation formats, namely, non-return-to-zero on–off keying and 4-ary pulse-amplitude modulation (4-PAM). The robustness of the higher order of AACC over 8-PAM and 16-PAM is also discussed.

High-mobility nickel (Ni)-doped zinc oxide thin-film transistors (NZO TFTs) have been successfully fabricated on flexible transparent plastic substrates at a low temperature. The devices all exhibited good electrical properties with small subthreshold swing (SS) and high saturation mobility fabricated under different sputtering pressure (0.4, 0.8, 1.2, 1.6 Pa) during channel deposition by RF magnetron sputtering were found. The electrical properties of NZO TFTs reached the best when the sputtering pressure was 1.6 Pa, with an SS of 89 mV/decade, a saturation mobility of 172 cm²·V⁻¹·s⁻¹, a drain current on/off ratio of 10⁸, and a positive threshold voltage of 2.36 V. The results show that Ni-doped ZnO is a promising candidate for flexible fully transparent displays.

The low timing jitter properties of a tunable single-pass optoelectronic frequency comb generator is investigated. The scheme is flexible in that both the repetition rate and centre frequency can be continuously tuned. When operated with 10 GHz comb spacing, the integrated residual pulse-to-pulse timing jitter is 11.35 fs (1 Hz–10 MHz) with no feedback stabilisation. The corresponding phase noise at 1 Hz offset from the photodetected 10 GHz carrier is −100 dBc/Hz.

By utilising an asymmetric metallisation scheme to fabricate an AlGaN-based solar-blind metal–semiconductor–metal photodetector, a zero-bias external quantum efficiency (EQE) of 24% for illumination at 240 nm wavelength from the substrate side was obtained. Moreover, an asymmetric bias-dependence of dark current and EQE is observed. The EQE saturates at about 38% in reverse direction, whereas in forward direction the presence of an internal gain mechanism is indicated.

Using the elasto-optic effect, increase of the frequency difference between the two orthogonally polarised modes, the so-called birefringence splitting, in standard single-mode oxide-confined AlGaAs-based vertical-cavity surface-emitting lasers is achieved to values beyond 250 GHz. A large birefringence is required for the generation of ultra-fast polarisation oscillations for potential future high-speed communication applications.

Expanding the photosensitivity spectrum of a single-junction GaAs-based solar cell to 1100 nm by using InGaAs hybrid quantum well dots (QWDs) multilayer media is reported. This nanostructure represents an In_0.3Ga_0.7As quantum wells with modulation of thickness and composition. Up to 15 QWD layers alternated with GaAs spacers can be inserted in an i-region of the GaAs p–i–n junction without impairing its crystal quality and quantum efficiency in spectral interval of GaAs absorption. The QWD layers are responsible for appearance of a longer wave spectral response (900–1100 nm). A photocurrent increment as high as 4.6 (5.2) mA/cm² for terrestrial (space) spectrum is demonstrated.

Geosynchronous synthetic aperture radar (GEO SAR) is susceptible to the ionosphere due to its ultra-long integration time and ultra-wide coverage, resulting in image defocusing and drift. A design of a validation experiment for analysing ionospheric impacts on GEO SAR focusing using global position system (GPS) signals is presented. The experiment consists of data acquisition and equivalent pre-processing of the GPS signals, along with the following contaminated GEO SAR signal modelling, imaging and evaluation. The key step is the pre-processing part for the equivalent transformation of the recorded data into the space–time frame of GEO SAR. Thereafter, the phase errors induced by the ionosphere in GEO SAR can be generated from the pre-processed GPS data and then incorporated into the simulated GEO SAR signals. Finally, the image focusing and impacts analysis are accomplished. In cases of the current inclined L-band GEO SAR system configuration, the existence of the ionosphere will induce image drifts but these can be corrected through image registration techniques. Another aspect is that the image can be well focused in range direction; however, the azimuth focusing is dependent on the second and higher derivatives of TEC, the thresholds would be determined based on the specific GEO SAR system configuration.

In skywave over-the-horizon radar (OTHR), transient interference is always suppressed with steps of interference localisation, corruption segment blanking and data restoration. The gapped data amplitude and phase estimation (GAPES) method has been proved to achieve good restoration performance. However, the large computational cost limits its application. A computationally efficient method for transient interference suppression is proposed. Quadratic smoothing is used to suppress transient interference instead of blanking the contaminated data. After that, the GAPES method only needs to restore the smoothed target echoes. Since quadratic smoothing can preserve most clutter echoes, the GAPES method will converge much faster. Experimental results demonstrated the effectiveness of the proposed method.

A novel technique for wideband radar cross-section (RCS) reduction of a microstrip antenna, with the radiating performance preserved at the same time, is proposed and investigated. The design is based on replacing the solid metal ground with the hybrid ground, which consists of suitable compact split-ring resonators. The antenna irradiated by both transverse-electric and transverse-magnetic polarised waves in the normal direction is analysed and investigated. Results show that RCS is dramatically reduced over a wide frequency range from 2 to 20 GHz. By comparison, remarkable RCS reduction of about 11 dB at 20 GHz can be obtained by the proposed strategy. Based on the acceptable effect on radiation features and prominent RCS reduction, the superior performance and reliability of the proposed design are demonstrated.

Unrelaxed neutral oxygen deficiency centres (ODCs) (V⁰ ODC II) in SiO₂ have been identified as the cause of random telegraph signals (RTSs) in highly scaled n-type metal–oxide–semiconductor field-effect transistors. Variable temperature RTS measurements were performed to extract trap capture cross-sections, capture activation energy, relaxation energy associated with the gate oxide defects, and the trap energy in the SiO₂ bandgap to determine the trap species and type. The results indicate that the electron is captured by a neutral ODC that is transformed into a negatively charged vacancy (V⁻). This is the first time V⁰ ODC II centre is confirmed to be the source of electron switching through RTS measurements defining four different trap characteristics.

An efficient resource allocation scheme is presented to maximise energy efficiency for multiuser orthogonal frequency division multiple systems with proportional rate constraints. For a determined subchannel assignment, an optimal energy-efficient power allocation algorithm has been derived through a bisection search method.

The optimal precoding design to maximise capacity of spatially correlated dual-hop non-regenerative multiple-input–multiple-output relaying systems is proposed. The correlated systems generally have additive coloured Gaussian noise at receive terminals and correlated data symbols at the source. Simulation results show that the capacity of the proposed precoding scheme is significantly higher than those of the existing schemes.

A QoS (quality of service) fairness-based slot allocation algorithm is proposed for sink stations in satellite-based sensor systems, which considers the statistical traffic rate and the instantaneous backlog info. Different from the traditional throughput fairness, the considered QoS fairness focuses on the long-term fairness, not the short-term fairness. The delay performance is taken as an example of QoS. With the help of stochastic gradient theory, a unique QoS-aware factor is deduced and used to direct the reallocation of the residual slots after basic allocation. Numerical simulations show that the proposed allocation could improve the long-term fairness in a large ratio, compared with the allocation for short-term fairness.