A 3-D non-uniform time step locally one-dimensional finite-difference time-domain (NUTS LOD-FDTD) method is presented. Unlike conventional and the other LOD-FDTD methods, NUTS LOD-FDTD method employs NUTS for different periods during the simulation. The update equation of NUTS LOD-FDTD method with arbitrary time step is given. Numerical results are provided to compare with conventional LOD-FDTD method and trade-off between efficiency and accuracy for different NUTSs.

Folded spherical helix (FSH) antennas using thin and wide copper strips instead of thick wires as a means of reducing the electrical size of ka is designed. It is found that FSH antennas made of copper strips show comparable radiation properties to FSH antennas made of thick wires. Furthermore, such designs can achieve even smaller size down to ka of 0.11 while keeping the radiation quality factors approaching the lower physical bound. An FSH antenna made of copper strips with ka of 0.21 is fabricated as a prototype using 3D printing technology. The measured resonance occurs at 307 MHz with the radiation efficiency of about 90%, verifying the simulated results and the proposed method of utilising copper strips in designing extremely small yet highly efficient volumetric folded helix antennas.

The impedance bandwidth (BW) improvement property of a self-resonant folded spherical helix electric dipole and a spherical split ring (SSR) magnetic dipole is compared when a negative reactance element is loaded on the parasitic resonator of the antennas. They have the same electrical size of ka = 0.38 and both approach each lower Q-bound. It is found that the impedance characteristic reacts more sensitively to the loaded element for the SSR antenna due to the inherent two-times higher Q of the magnetic antenna. It is also found that a conventional non-Foster circuit composed of a series negative inductor and a capacitor cannot be used for wide impedance matching of the SSR antenna due to the high sensitivity. Subsequently, a circuit topology that enables more precise matching of the reactive values for widened impedance matching in the high-Q SSR antenna and satisfies the stable operating condition is proposed. It is observed that the improvement ratio of the BW is limited in practical situation. Nevertheless, the measured reflection coefficient of the 3D-printed SSR antenna loaded with the designed active matching circuit shows two-times wider impedance BW than passive matching, verifying the simulated result from the circuit and electromagnetic (EM) simulations.

A single ring two-port multiple-input–multiple-output (MIMO) antenna is proposed. Unlike conventional techniques such as space decoupling structures or decoupling networks, two signal paths in this mono-radiator antenna cancel out each other to achieve good isolation. A MIMO antenna covering 33–41 bands of time division duplex long-term evolution (LTE) cross 1.85–2.62 GHz is designed, fabricated and measured in this Letter. Simulations are confirmed by the experimental results. The details of design approach and process are introduced.

A novel wideband dual-mode planar endfire antenna with circular polarisation is presented. The antenna consists of a semicircular disk radiator and a concentrically, stub-loaded annular sector dipole. A pair of stubs is symmetrically loaded along the annular sector dipole so that the second odd-order dipole mode is excited and moved downwards to the fundamental mode. Thus, a wideband, planar endfire circularly polarised antenna with dual-resonant characteristic is achieved. It is shown that the prototype has an impedance bandwidth of more than 24.8%, and a 3 dB axial ratio bandwidth of 18.8%.

An extended Wang neural network (EWNN) is proposed to solve online a set of linear equations. Such EWNN is possessing the general nonlinear model form with redundant parts, to face existence of nonlinearity phenomena in circuit implementation of Wang neural network (WNN). Furthermore, two types of nonlinear activation are proposed for EWNN aiming to improve the convergence of the WNN. Illustrative results verify the proposed EWNN for online solving linear equations.

Inversion layer technique can increase the centre frequency (f ₀) and the −6 dB bandwidth of the ultrasound transducer without reducing thickness of active material. However, it suffers from the reduced −6 dB bandwidth when the centre frequency is increased more than 1.5f ₀. A modified inversion technique is proposed and applied it to high-frequency intravascular ultrasound (IVUS) transducer. The performance was computationally verified by using finite element method (FEM) simulation. The proposed IVUS transducer consists of two sub-elements with each aperture size of 0.6 × 0.6 mm, and each sub-element has different types of inversion layer using PMN-PT and PZT-5H. The initial thickness of overall active material was designed for 40 MHz centre frequency and the FEM simulation showed that the centre frequency was increased by about 1.8f ₀ maintaining the −6 dB fractional bandwidth compared with the conventional transducer.

A miniaturised implantable antenna with dual-band operation the Medical Implant Communications Service (MICS) (402–405 MHz) and Industrial, Scientific, and Medical (ISM) (2400.0–2483.5 MHz) bands is presented. The size of the proposed antenna is 31.5 mm³ (8.75 mm × 7.2 mm × 0.5 mm) which is the smallest size compared to previous implantable antennas. A serpentine-shaped radiating patch and open-end slot placed on the ground plane are used for miniaturisation. The performance of the antenna was evaluated from measurements and is based on good agreement with simulations.

A dual-modulus prescaler based on true single-phase clock D flip-flop is presented. Instead of using the conventional digital logic circuit, the pass transistor logic circuit is applied to reduce the number of the transistors and the power of the divider. By adding only two additional transistors, a dual-modulus operation is achieved. Implemented in a standard 180 nm CMOS process, the proposed divide-by 6/7/8 prescaler based on the proposed dual-modulus one achieves an operating frequency of 4.2 GHz with a measured power consumption of 0.81 mW for a 1.2 V supply.

A memristive Chua's circuit regarded as a paradigm is reconsidered to exhibit its extreme multistability in this Letter. Memristor initial state-dependent dynamics is analysed and the coexistence of infinitely many attractors related to memristor initial states is revealed by numerical simulations and circuit simulations. The dynamical behaviour just reflects the emergence of extreme multistability in the memristive circuit.

Several dynamic behaviours can simultaneously coexist in the operation of a nonlinear electronic circuit. However, some of these dynamics are difficult to be observed due to their sensibility to initial conditions. The use of describing functions to determine the coexistence of multiple dynamics in the operation of an electronic circuit with nonlinear feedback is proposed.

Novel two-transistor embedded memory – floating body gate cell – is implemented on planar SOI CMOS technology without adding extra masks. Since channel current is designed for memory cell write operations, this cell demonstrates ultra-fast write speed which is comparable with static RAM cell. The decoupled write and read structure ensures small operation power consumption and avoid false read. The low operation voltages of this cell lead to the excellent endurance performance. In addition, retention time is greatly enhanced due to the gate-to-drain underlap design.

An improved dual-capacitive arrays DAC (IDCA-DAC) architecture for successive approximation register ADC is presented. By using a switching-back switching scheme and a third reference voltage for the last bit conversion, the proposed IDCA-DAC architecture achieves 42% savings in average switching energy and 47.1% reduction in area compared with the original DCA-DAC architecture. Additionally, it can achieve 99.3% savings in average switching energy and 71.9% reduction in area compared with the conventional method.

A second-order charge-sampling structure is proposed which enables the implementation of complex conjugate poles. The proposed structure avoids the use of any extra active element and reaches the desired response by modifying the switching scheme and inserting a feedback mechanism in the operation of the filter. Simulation results show that the proposed structure is capable of implementing a Butterworth filter.

A brief example of application of recently introduced concept of electronically controllable CMOS voltage differencing current conveyor of second generation in electronically tunable second-order oscillator is shown. On-chip experimental results confirm theory and design intentions.

High-efficiency video coding (HEVC) outperforms H.264/AVC by providing a bitrate reduction of about 50% while having almost the same perceptual quality. It adopts more flexible partitioning in motion estimation (ME) which gains higher coding efficiency at a cost of increased coding complexity. Depth maps in the emerging multi-view plus depth videos are exploited to adjust the search range in ME for HEVC complexity reduction. With the aid of depth intensity variations among neighbouring blocks, the proposed algorithm derives weights to the neighbouring blocks and establishes an adaptive search range according to the weighted sum of the motion vectors from the neighbouring blocks. Compared with the fast test zone (TZS) search in HEVC, the proposed algorithm saves 65% coding time in ME on average with insignificant rate-distortion degradation.

Depth from defocus (DFD) technique calculates the blur amount in images considering that the depth and defocus blur are related to each other. Existing DFD methods generally compute the blur at edge locations and solve an optimisation problem to propagate the blur from edges to all image pixels. Solving the pixel-based optimisation problem is time-consuming, posing the performance bottleneck. Moreover, the generated depth maps are not consistent in textured areas and the blur estimation may be incorrect in the regions with soft shadows. We address these problems by proposing a superpixel-based blur estimation method. Experimental results show that the proposed method is not only faster than pixel-based blur estimation, but also can improve depth data in textured regions and soft shadows.

A low complexity but effective method in image and video detail enhancement – in-place similarity is proposed. In-place similarity is the statistic result of different kinds of nature image patches when similarity measurement is done in super resolution research, finding that patches in the upscale image have a good match around its original location in the corresponding lower ones. On the basis of the in-place similarity, a simple assumption is made that detail layer is the high-frequency component of twice search and patch match result. Then, images under the framework of super resolution techniques are enhanced. Unlike many current algorithms that need to adjust parameters for different images to acquire best outputs the approach mentioned here is adaptive. Moreover, many algorithms have outcome with intensity and contrast change, the algorithm proposed here can prevent images from over enhancement with obvious nature looking effect. Extensive experiments demonstrate that the algorithm mentioned here is simple with robust good performance both subjectively and objectively. It is particularly useful for practical applications and easily hardware implemented in FPGA, where input images require diverse detail textures.

Bidirectional reflectance distribution function (BRDF) estimation is one of the fundamental problems in computer graphics, computer vision and related fields, yet most work relies on conventional local minimisation. Two global optimisation methods for estimating the anisotropic Ward BRDF model and its variants under an L ₂-cost function are introduced. The presented algorithms are based on branch-and-bound search which guarantees to provide globally optimal solutions. In contrast to the traditional local optimisation, the methods mentioned here are not affected by the setting of initial BRDF parameter values. Experiments have been carried out to validate the presented methods using anisotropic materials from the UTIA BRDF database.

A hard-decision iterative decoding scheme for the Davey–MacKay construction with symbol-level inner decoder is presented. In this scheme, the synchronisation capability of the inner decoder is enhanced by replacing the equal priori probabilities for symbols by the accurate hard-decision feedback information from the outer decoder. Furthermore, the error-correcting capability of the outer decoder is improved and the hard-decision feedback information is updated. Simulation results show that the presented scheme significantly improves the performance and saves the computational time with little performance loss.

A novel switching circuitry for fully differential millimetre-wave applications is presented. The structure of the proposed switching architecture includes 90 nm CMOS transistors and a distributed-active-transformer. To minimise the chip size, an innovative topology based on lumped components is investigated, resulting a total size of <0.025 mm². The performance of the designed switch is validated within a dual-mode local oscillator-distribution circuitry of a V-band transceiver.

A switchable wideband reflective metamaterial polarisation rotator based on plasmon resonances has been presented. The proposed design comprises of periodically arranged split rings on which PIN diodes are mounted to demonstrate switching operation. By controlling the bias voltage of the PIN diodes, the polarisation of the reflected wave can be controlled over a broad frequency range. The simulated result shows that over 90% cross-polarised reflection is observed in 2.42–3.52 GHz under OFF state, whereas co-polarised reflection is exhibited under ON state. Polarisation rotation has also been explained using two plasmon resonances in terms of electromagnetic responses. Finally the structure has been fabricated, whose measured results agree well with simulated responses, thus validating the concept behind switchable polarisation rotator.

A novel method of bonding surface mount components to screen-printed fabric circuits is proposed. The method is achieved by applying the same technique of screen-printing conductive ink patterns of fabric circuit for printing a layer of low-temperature solder paste onto uncured conductive ink pads, and placing the circuit with mounted components inside an oven for a one-step curing process. A screen-printing frame with 42 μm mesh opening is used for both conductive ink and solder paste layers. Thus, a 0.2 mm resolution for printed traces and spaces is achieved, which meets the demanding geometrical requirements of most integrated circuit chips with fine-pitch and tiny footprints. Measurement results show the printed tracks after curing exhibit excellent conductivity while retaining good flexibility. Examination under the microscope shows the components are well-bonded to their on-fabric pads. The proposed method for surface component mounting and one-step curing process for screen-printed fabric circuits can potentially streamline and expedite the twin process of circuit fabrication and component mounting for future smart textile applications.

A half-mode substrate integrated waveguide (HMSIW) bandpass filter loaded with modified complementary split-ring resonators (CSRRs) is presented. The modified CSRR, realised by utilising horizontal–asymmetrical stepped-impedance (SI) structure in the conventional CSRR, can enhance the equivalent capacitance and inductance of the CSRR and consequently shift the resonant frequency downwards. Therefore, for the same operation frequency, the SICSRR can occupy smaller size than the conventional CSRR, which provides a promising method to achieve size-reduction in microwave circuits. To verify the effectiveness of miniaturisation introduced from the proposed HMSIW-SICSRR resonator, a two-pole bandpass filter is implemented. Moreover, the measured results are in good agreement with the simulated ones.

An optimal bit-and-power allocation (BPA) algorithm for visible light communication-orthogonal frequency-division multiplexing (VLC-OFDM) system is proposed and demonstrated. It is such notable that the signal noise in VLC-OFDM system is not addition white Gaussian noise and its power changes with the received signal power through experimental measurement. Thus the optimal Hughes–Hartog BPA algorithm based on variable noise power is presented. Compared with the traditional BPA algorithm, the experiment indicates that the proposed optimal BPA algorithm can achieve about 15.9% improvement of data rate for VLC-OFDM system.

The stress-induced breakdown characteristics in magnesium oxide (MgO)-based magnetic tunnel junctions (MTJs) including an inserted Mg layer 0.5 nm thick is investigated, above or below an MgO tunnel barrier. Regardless of the insertion position, the inserted layer suppressed the time-dependent dielectric breakdown (TDDB) observed in the case of electron tunnelling into the Mg-inserted interface. This indicates that the Mg insertion significantly suppressed trap site formation at the anode-side barrier/electrode interface. The improvement of TDDB by the Mg insertion was confirmed through constant voltage stress experiments. Therefore, interface modification by means of inserting an ultra-thin metallic layer is highly effective in improving the reliability of an MTJ tunnel barrier for practical applications.

The fabrication of a flexible electronic system which consists of a flexible polyimide (PI) substrate and a rigid silicon thin chip is demonstrated. The 50 μm thick rigid silicon chips are bonded to the PI substrate using a die attach film (20 μm) and electrically interconnected to an adjacent metal pad which is deposited on the PI substrate and silicon chips. A 3D step covered interconnection consisted of silver nanoparticles and polymer are achieved by using an electrohydrodynamic system. The mechanical and electrical characteristic was investigated by performing a Kelvin resistance measurement on the jetted lines while the substrate was subjected to bending modes of varying diameters.

A comparative analysis of the properties of vertical light-emitting diodes (VLEDs) fabricated using different surface treatment schemes was conducted. Compared with a conventional VLED, a light output power increase of 24.8% for an injection current of 350 mA was achieved by an n-type-gallium nitride (GaN)-based VLED with a thin undoped GaN layer and microsized protrusions produced by krypton fluoride laser irradiation using an energy density of 600 mJ/cm². This was accomplished without noticeable degradation of the electrical properties of the device. Further, potassium hydroxide wet etching of the VLED surface increased the light output power gain to 47.3% for the same injection current.

A novel polarisation beam splitter (PBS) with a high extinction ratio in a broad band has been proposed and realised by cascading a bent directional coupler and a low-loss transverse-magnetic (TM)-pass polariser based on a subwavelength grating. The realised PBS has a footprint as small as 3 × 22.5 μm, a high extinction ratio (>20 dB) and a low excess loss (∼1 dB) are obtained over a broad band of ∼60 nm for both transverse-electric and TM polarisations.

An analogue–digital circuit for thermal control of the resonance frequency of optical ring resonators is suggested. This approach is characterised by a low power dissipation of 209 µW and needs only 2800 µm² of chip area in 0.16 µm CMOS technology. Therefore, it is well appropriate to be used to control each of the cells of large optical switch matrices. Compared with a purely analogue heater control approach the power dissipation is reduced to 7%.

A highly nonlinear W-type erbium-doped fibre with high dopant concentration made of three lead silicate glass materials (Schott SF57, LLF1 and SF6) is presented. The fibre is designed to have a flat anomalous dispersion at ∼1.55 μm, which is critical for enhancing parametric wavelength conversion. With this fibre, signals at ∼ 1.55 μm are successfully amplified in continuous-wave and pico-second-pulsed formats with a 1.48 μm pump source. In particular, spectral broadening of over 350 nm together with strong degenerate four-wave-mixing effects is demonstrated. The authors highlight that this novel fibre combines, for first time to their knowledge, the benefits of highly doped active fibres with those of dispersion/nonlinearity-engineered fibres.

This Letter reports a lead (II) ion sensor composed of a tapered multimode fibre (MMF) coated with chitosan. The diameter and length of the tapered fibre region is 20 µm and 1 cm, respectively. Tapered MMF is used because it is more flexible and compact compared with plastic clad silica. Chitosan is chosen as the sensing layer because it can absorb heavy metal ions due to its unique properties as polyelectrolyte and chelating agent. The refractive index of the chitosan layer changes as it absorbs lead (II) ion and this leads to an increase in the absorbance measured by the spectrophotometer. This is tested by immersing the sensor in a solution that contains lead (II) ion with different concentrations, from 0.2 to 1 ppm. The results show that the absorbance measured by the spectrophotometer increased as the concentration of the lead (II) ion increased. The sensor's average sensitivity is measured as 40.554 abs/ppm. This sensor is simpler compared with surface plasmon resonance-based lead (II) ion sensor because it does not require metal coating prior to the sensing layer coating.

A new approach is presented in order to determine copper losses from measurements or simulations of planar inductors. Copper losses are modelled with a series resistance r _(f) that depends on the frequency. Measured or simulated admittances of the inductors are used to determine the model parameters of the inductors. The resistance r _(f) is determined in three frequency domains: (i) at very low frequencies (or DC); (ii) at low frequencies (capacitive couplings are negligible); (iii) at resonant frequencies, i.e. high frequencies. The presented approach is different as those encountered in literature and allows the series resistance from measured or simulated S_ij parameters to be determined.

A high efficiency single-stage switching amplifier based on high-frequency link inverter is proposed. Bipolar phase shift modulation is adopted in the proposed switching amplifier. Auxiliary circuits including freewheeling diodes, clamp capacitors and active clamp bridge (ACB) are added to provide a current flowing path for the inductor current during the dead time of the bidirectional switches. The ACB is synchronised with primary-side switches to absorb the leakage inductor energy and clamp the secondary-side transformer voltages. With single-stage conversion and leakage inductor energy recycle, high efficiency of the proposed switching amplifier is achieved. A 5 V/25 W prototype is presented to verify the analysis results of the proposed switching amplifier.

An integrated balancing transformer is proposed for a high frequency parallel AC-LED drive circuit using a LLC resonant converter. The design and implementation of the integrated balancing transformer are presented. The experimental results for the converter with the integrated transformer are provided to show the validity of the proposed technique.

A high-side load switch is presented for applications that require multiplexing the supply terminal to multiple voltage sources that are different, unknown, and/or dynamic. The design incorporates p-channel switches with adaptive n-well biasing to ensure that no junction is forward biased. Measured results are presented from a circuit that has been fabricated in a standard 0.35 μm CMOS process. The circuit is also demonstrated in an application of programming non-volatile memory.

In highly sloped terrain, the interferometric phase gradients of the adjacent pixels are more than π. Therefore, the conventional phase unwrapping methods cannot correctly obtain the absolute phase. To solve this problem, a phase unwrapping method based on multi-frequency interferometric synthetic aperture radar (MF-InSAR) is proposed. In this method, a joint probability density function based on MF-InSAR is established to estimate the phase gradient in highly sloped terrain, and thanks to the small changes of the phase gradient differences between adjacent pixels, an adjacent set of the phase gradient is further proposed to improve the robust of this algorithm. After the reconstruction of phase gradients, the branch-cut phase unwrapping method is used to obtain the ultimate absolute interferometric phase. Computer simulation proves the proposed method can correctly obtain the absolute interferometric phase in highly sloped terrain, which makes further efforts to acquire a better reconstruction accuracy of DEM.

A near field targets localisation method using parallel factor (PARAFAC) decomposition with a bistatic MIMO system is presented. PARAFAC has the interesting properties of uniqueness in the decomposition of tensor constructed from the received signal and automatic pairing of emitter and receiver directional vectors of a bistatic MIMO radar, which allow the estimation of the parameters of each source individually. Furthermore, the proposed method provides very good estimation performance, because it does not require the approximation of the directional vectors contrary to the most existing techniques.

A Doppler radar based microphone is similar to an acoustic microphone except that it detects vibrations rather than sound. A microphone with a logarithmic square-law detector for monitoring honeybees is described. The performance of a logarithmic square-law detector is compared with a more conventional zero-intermediate frequency (IF), direct down conversion receiver. It is shown that unavoidable coupling between the transmitter and receiver as well as stationary clutter have the same effect in the square-law receiver as the local oscillator signal and mixer have in the direct down conversion receiver. The theoretical output of either configuration is formally the same. Both receiver designs were implemented in a low-cost, low-power radar microphone operating in the unlicensed 5.8 GHz industrial, scientific and medical band. The use of a logarithmic square-law detector resulted in a significant simplification of the radar microphone design as well as a 10 dB improvement in signal-to-noise ratio.

Real-time differential (RTD) positioning can achieve a real-time metre-lever positioning based on the differential correction data from a nearby reference station. However, the time-delay of correction data is generally unavoidable when receiving it from the reference station in real-time. Multi-global navigation satellite system combined positioning has the advantages of improving the positioning availability, continuity and accuracy; however, it remains unknown whether it can effectively overcome the time-delay of differential correction in RTD positioning. This Letter focus on the tolerance of time-delay for differential correction in BeiDou global navigation satellite system (BDS) and global positioning system (GPS) combined RTD positioning. Based on a set of static data and a set of dynamic data, the performances of BDS standalone, GPS standalone and BDS and GPS combined RTD positioning by 15 kinds of time-delay of the differential correction data were studied and compared. The numerical results show that, BDS and GPS combined RTD positioning can tolerate longer time-delay of the corrections than BDS or GPS standalone positioning for the same requested positioning accuracy.

A moving train was imaged by a Ka-band interferometric (In) radar of 2 GHz bandwidth working at the dechirping model. Two radar images from two channels as well as their In image were obtained after estimating the speed and direction of the train. The plots of In phases (InPhes) along the train exhibits periodic fluctuations, which are resulted from the back and forth motions of both locomotive and other compartments when the trains are moving ahead, and this can be attributed to the micro-motion (m-M) of the train. As far as it is known, this is the first time to experimentally demonstrate m-M detection through InPhes.

On the basis of the intrinsic sparsity of sea clutter in shipborne high-frequency surface wave radar, a joint sparse recovery-based space–time adaptive processing (STAP) technique is proposed to suppress the spread sea clutter. In the proposed method, the partial clutter subspace is first estimated with training samples from only one range cell. Afterwards, a generalised matching pursuit method using the oblique projector is derived to recover the complementary clutter subspace. The sea clutter suppressor is finally constructed by STAP. According to experimental results, the proposed method can obtain a more accurate estimation of sea clutter spectrum and thus achieve better suppression performance.

We successfully fabricated the fully transparent Al–Sn–Zn–O thin-film transistors (ATZO TFTs) on glass by RF magnetron sputtering, and then the electrical performances of the ATZO TFTs are optimized with double-channel structures. The double-channel structures are formed by using bilayer ATZO films fabricated successively with different oxygen partial pressure in the sputtering process. The operation mechanism for double-channel structures were clarified. Owing to the double-channel structure, the ATZO TFT demonstrates excellent electrical performances, including a high ON/OFF current ratio (I _on /I _off) of 1.1 × 10⁸, a steep threshold swing SS of 266.3 mV/decade, a superior saturation mobility μ _sat of 134.1 cm²/Vs, and a threshold voltage V _T of 1.2 V.

Two strategies are introduced herein to improve current drivability of tunnel field-effect transistors (TFETs). First, gate-to-channel coupling is increased by ∼22% in terms of effective gate capacitance (C _ox) with the help of hemi-cylindrical device architecture. A novel iterative corner rounding process was developed for highly reliable gate dielectric formation without field crowding at sharp corner. The second approach is change channel orientation. The 〈100〉 and 〈110〉 oriented-channel TFETs were fabricated on (100) silicon surface. Consequently, both planar and hemi-cylindrical structures show consistent results that 〈110〉 is far better than 〈100〉 for higher on-current (I _on). With these two structural improvements, the 〈110〉 hemi-cylindrical TFET shows ∼30× higher I _on than that of the control (i.e. 〈100〉 planar TFET).

A multilayer (ML)-based resistive RAM (RRAM) is proposed to achieve the performance of logic and memory application. Multilevel resistive switching characteristic of ML-based RRAM is demonstrated in aspect of data storage application. An innovative method is proposed to simultaneously achieve quaternary addition and data storage application, resistance serves as physical state variable instead of voltage or charge in the computing system. An instance of quaternary non-volatile addition is demonstrated in an experiment, which shows the ML-based RRAM is promising for non-volatile computational application as well as higher integration capability.

The features of just noticeable difference (JND) in human spatial perception are often used to remove the perceptual redundancy of spatial parameters in 3D multichannel audio. However in previous spatial parameters quantisation schemes, JND data are not effectively utilised resulting in either perceptual distortions of spatial images, or wastes of coding bitrates. It is proposed to effectively utilise azimuthal JND to design the quantisation codebook of spatial parameters azimuths, which divides the full circle of 360° to adjacent quantisation intervals, and set the endpoints of quantisation intervals according to JND of quantisation value. With presented scheme, quantisation codebook size has decreased 13 and 57%, respectively, compared with MPEG Surround and SLQP, as well as ignorable perceptual distortion induced.

A divide-by-3 frequency divider based on heterodyne injection-locking technique is presented. An injection-locking LC oscillator with resistive feedback and differential signal injection is implemented to achieve higher injection efficiency over a wider range. Moreover, by optimisation of inter-stage loading as well as filter-less design, only one inductor is used in this design to save silicon area. Implemented in a standard 0.18 µm CMOS process, the proposed divide-by-3 divider has a locking range from 17 to 19.1 GHz without extra frequency tuning and consumes 10 mA from a 1.8 V supply voltage.

Most of the range-free localisation techniques compute the position of a node through multilateration based on distances between nodes. The least squares estimation (LSE) based multilateration is used to determine the position of a node – which is optimal estimation method if all the variances of distance errors are equal. In range-free localisation, the distances to the anchors are usually estimated by multiplying the hop count and the hop length. Since the estimated hop length is error prone, the distance estimation error and the error variance usually increase as the hop count increases. If the conventional multilateration algorithm, which gives the same weights to the estimated distances, is applied as it is, suboptimal result will be achieved because the variances of the distance estimation errors cannot be equal. In this Letter, we analyse the reliability of estimated distance and propose a novel LSE based multilateration which considers the reliability of estimated distance. Simulation results show that the proposed method improves performance compared with the existing methods.

Precise estimation of inter-sensor distances is essential for reliable localisation in Internet of Things sensor networks. A cost-effective, scalable, asynchronous solution to estimate inter-sensor distances based solely on measurements of distances to a moving object is proposed. More specifically, the proposed solution estimates uncharted distances using trigonometry and processes these estimated distances with a distributed weighted multi-dimensional scaling algorithm for more precise localisation of sensors. It is demonstrated that the proposed solution achieves the localisation error of 4.8–33.9 cm when the measurement errors of sensor devices are in the range of 5–40%.

Beamforming (BF) transmission for broadband massive MIMO systems in frequency-selective Ricean fading environments is investigated. By treating the scattered component as interference in order to reduce the heavy overhead for estimating channel state information, a transmit/receive BF scheme only based on the specular component is presented. It is shown by analytical results that the ergodic achievable rate for the scenario with doublyended spatial correlation can be asymptotically equal to that of the one without correlation when the transmit power is scaled down by the multiplication of the two numbers of antennas at the transmitter and the receiver. It is also shown by simulation results that different from the traditional point of view, spatial correlation can improve the rate performance of the proposed BF scheme.