Multi-functional antennas are very important in modern wireless communication systems. A novel dual-band frequency-reconfigurable antenna with pattern diversity is designed. The antenna consists of a composite right/left-handed transmission line inspired structure, which works at the zeroth-order mode (0) around 2.5 GHz and at the first-order mode (1) around 3.5 GHz. The 0 mode results in an omni-directional pattern and the 1 mode results in a broadside pattern. By adding varactors to continuously reconfigure the structure, the antenna is able to tune its lower resonant frequency from 2.36 to 2.60 GHz, which fully covers the 2.4–2.5 GHz ISM band, and tune its higher resonant frequency from 3.34 to 3.73 GHz, which fully covers the 3.4–3.7 GHz WiMAX band. Simulations and experiments match well, proving the validity of this new concept.

A compact self-isolated multiple-input–multiple-output (MIMO) antenna system is proposed. It is based on a quarter-mode substrate integrated waveguide (SIW) cavity where two edges are shorted by metallic vias and the other edges are open. A narrow slot with a rotated angle of 45° is utilised to divide the single cavity into two equal eighth-mode sub-cavities. These sub-cavities are fed by two coaxial ports to construct the proposed MIMO antenna system. A self-isolated mechanism that cavity energy inputted by one port is radiated into free space rather than transmitted to the other port is first observed. High isolation between two antenna elements can be obtained without additional isolating or decoupling networks although they are contacted with each other. A fabricated prototype operating at 3.5 GHz achieves the measured port isolation of higher than 18.5 dB, envelope correlation coefficient (ECC) of less than 0.04, the peak gain of 4.6 dBi, and maximum efficiency of 79.2%. With advantages of compact size, high isolation, low ECC, and good radiation performance, the proposed self-isolated MIMO antenna system is a good candidate for wireless communication.

A compact four-element UWB MIMO antenna is proposed and its characteristics are investigated. The proposed antenna has a hexagon molecule-shaped fractal structure as its radiating element. The antenna elements are placed orthogonally to each other in order to yield good isolation, without any additional decoupling structure. The band rejection in WLAN band is attained by etching a C-shaped slot on each radiating element. The antenna has a compact size of 40 × 40 × 1.6 mm³ and exhibits stable omni-directional radiation pattern. In addition, it shows an acceptable impedance bandwidth (S ₁₁ < −10 dB) in the range 2.4–10.6 GHz with an isolation better than −20 dB over the UWB range. The performance of the proposed antenna is calculated in terms of antenna characteristics and MIMO parameters. The isolation level in-between the elements is found to be reasonable for MIMO application. The simulated and measured results are in good accord which makes the antenna a suitable candidate for massive MIMO and high-density packaging applications.

Also according to the recent Directive by the European Parliament and the Council of the European Union, there is a need for a simple but reliable device to measure exposure to the magnetic field (MF) of magnetic resonance (MR) workers, especially given the continuing development of scanner using a higher static MF. Here the authors describe a novel device for assessing personal exposure to the MF. The time recording system can measure the static MF value at which the operator is exposed to and the field gradients due to his/her movements in the MR environment. The device does not require any cable for data management and transmission and it is able to record the instantaneous and cumulative exposition dose. The device can be used in all MR environments, for clinical or industrial applications.

This Letter reports a 25.80–30.84 GHz wideband millimetre-wave VCO in 65-nm CMOS technology. A 4-bit binary-weighted capacitor bank is used to extend the tuning range. The PMOS current source, as well as the LC noise filtering network, are used to improve the phase-noise performance. Occupying a core area of 0.056 mm², the implemented VCO core consumes 4.1 mW at a 1 V supply. At 26 GHz, it achieves a phase noise of −102.8 dBc/Hz at 1 MHz offset.

Recently an extremely linear voltage-controlled ring-oscillator for use in VCO-ADCs was proposed by Babaie-Fishani and Rombouts. In this current Letter, a circuit-level technique to improve the bandwidth and noise performance of such a linear VCO is proposed. The key element is the modified delay cell, where the traditional cross-coupled inverters are modified into ‘feed-forward’ coupling inverters that pre-charge the subsequent elements in the ring. The resulting circuit maintains the excellent linearity, but has greatly enhanced bandwidth (up to 3 times higher) and considerably reduced power for the same circuit noise level (up to 2.5 times lower).

This Letter presents a digital low-dropout regulator (LDO) that achieves fast transient response with a flash ADC-based parallel comparison. A non-linear decoder is employed in the flash ADC to further enhance the transient response. In the design of the power switch array, an exponential-ratio array (ERA) is adopted for high load driving capacity. The proposed digital LDO implemented in 65-nm CMOS achieves a load range of 0.04–82.7 mA when the input voltage and the output voltage are 1.0 and 0.9 V, respectively. The digital LDO achieves a settling time of 6 μs with a load step of 48 mA, exhibiting the state-of-the-art normalised settling time for the clock frequency of 1 MHz.

Switching converters are widely used. Electromagnetic compatibility (EMC) is an important index in the design process. At present, there are three problems: (i) EMC noise test is usually carried out after the prototype design is completed. This method, which is not tested at the design stage, increases the design cost to some extent; (ii) the complexity and limitations of the test environment lead to low confidence of test results; (iii) traditional modelling methods ignore many factors such as power supply feeders. To solve the above problems, an electromagnetic interference conduction model is proposed in this Letter. After experimental verification, it is found that the test results of this model have a high matching degree with the measurement results of standard EMC equipment.

The authors introduce a novel approach for swarm reinforcement learning that extends the standard Q-learning to multi-agent systems. State-of-the-art methods implement a knowledge sharing mechanism between the agents that is triggered by the episodes succession. This causes an intrinsic limit in the convergence speed of the algorithms. They overcame this issue by developing a Q-learning real-time swarm algorithm (Q-RTS), which is iteration-based and suitable for real-time systems. Q-RTS was tested in different environments and compared to other related methods in the literature. They obtained positive results in terms of learning time and scalability, i.e. achieving a speed-up factor of at least 1.49 with respect to standard Q-learning. Moreover, Q-RTS shows enhanced learning performance as the environments complexity increases.

The finite-difference methods (FDMs) suffer from the notorious numerical dispersion problem, especially when large simulation time steps and wideband pulses are involved. In this Letter, an approach to obtain optimised numerical dispersion for FDMs based on the adaptive wind driven optimisation (AWDO) is proposed. Numerical dispersion errors are analysed and minimised through the AWDO, then the authors could obtain the updating coefficients for the FDMs. The 2D case is exemplified to demonstrate the procedures in detail. Serval numerical experiments are carried out to validate the accuracy of the proposed approach.

Vertical projection histograms are an efficient shape representation for 2D binary silhouettes and have been widely used in pedestrian localisation for video surveillance. The weakness of this method is that it is not invariant to rotation. In this Letter, a generalised vertical projection histogram is proposed to solve this problem, in which the homology transformations of the foreground silhouettes, for a set of parallel planes, are warped to, and accumulated, in the original foreground map. Then a method, similar to the vertical projection histogram, is carried out to localise the pedestrians in the foreground silhouettes. The algorithm applies an integrated approach using both image projection and geometric projection. Its value is demonstrated in a case study on pedestrian localisation with cast shadows.

Part misalignment of the human body caused by complex variations in viewpoint and pose makes person re-identification be a fundamental challenging task. In this Letter, the authors propose a new Part-Aligned Network (PAN) stacked by Part-Aligned Module (PAM) to learn the effective representation for person re-identification. Specifically, PAN is constructed by stacking several PAMs which is simultaneously comprised of an appearance branch and a part branch. The appearance branch performs spatial modelling while the part branch captures human part information simultaneously. Appearance features and part features are then combined to explore the implicit complementary advantages and these two branches are learned in a mutually reinforcing way. Extensive experiments are conducted on the three most popular datasets: Market-1501, DukeMTMC-reID, and CUHK03. The experimental results demonstrate that the proposed PAN achieves superior performance to the existing state-of-the-art methods.

Generative models such as variational autoencoder (VAE) and generative adversarial network (GAN) have been widely applied to many areas including image synthesis and voice generation. However, they have some problems that VAE makes blur images and GAN is difficult to learn due to mode collapsing. A novel generative model is proposed using a self-organising map (SOM) termed a self-organising generative network (SOGN). In the SOGN, training images are first mapped to SOM and then the output space of SOM is transformed into 2D vector spaces. These vector values are used as latent vectors to train the generative network such as artificial neural networks or convolutional neural networks. Experimental results with MNIST and CIFAR-10 datasets showed that their generative model was easy to train without mode collapsing and made more clean images than VAE. It was also confirmed that the manifold is well-observed without generating by the average effect of multiple images.

Sequence pairs with good correlation can be applied in many practical systems. In this Letter, the authors construct a class of p-ary sequence pairs of period by interleaving two p-ary sequences derived from Fermat quotient, where p is an odd prime. In the authors' construction, sequence pairs have almost perfect correlation.

In this Letter, a list decoding algorithm is proposed for the semi-random block-oriented convolutional code (SRBO-CC). To decode each sub-frame, a list of candidates are serially computed until finding a qualified one or achieving maximum list, where the correctness of the decoding candidate is checked by a statistical threshold. Simulation results show that SRBO-CC with list decoding is competitive with polar codes and that the performance-complexity tradeoffs can be achieved by adjusting the statistical threshold.

To improve the performance and signal integrity for printed circuit board designs, considering differential pairs in the escape routing is an important issue. Previous works use two-stage approaches, meeting points determining followed by differential pairs connecting, to generate a routing solution. However, existing approaches may produce either non-length-matching solutions or longer length-matching solutions. In this Letter, the authors propose an escape routing model with one-stage optimisation to simultaneously determine median points and route length-matching paths for all differential pairs. Experimental results show that their approach can significantly reduce maximum and average differential-pair skews with 100% length-matching results.

A non-alternating (NA) form of K-means is proposed to improve the performance for the most fundamental, yet highly non-convex clustering problem. The motivation of this Letter is that the non-convex nature of K-means can be better handled by stochastic optimisation. However, the alternating update of the Lloyd's algorithm prohibits an effective stochastic optimisation. In order to fully realise the idea, a probabilistic representation is provided for the solution space of K-means, which leads to a simple yet efficient NA update. Experiments show that the proposed method outperforms the existing variants, especially for large numbers of clusters, with reasonable time complexity.

A band-stop filter based on spoof surface plasmon polaritons for microwave applications is proposed, which can realise high-efficiency band rejections by etching gradient semi-circular grooves on both sides of the middle stripline of coplanar waveguide. The operating bandwidth and frequency of band-stop can be controlled by changing the radii of semi-circular holes and the length of unit cell, respectively. The operating principle of the designed band-stop filter is explained by the dispersion curves and electric field distributions. Parametric studies are also conducted to analyse the effects of key parameters on the performance of the band-stop filter, which is useful for practical designs. A prototype of the band-stop filter is fabricated and measured. The good agreements between the simulated and measured results validate the proposed design, which indicate that the presented band-stop filter can be potentially useful in many microwave applications.

A novel reflectionless triple-band bandpass filter (RTBPF) with compact structure and high controllability is proposed by combing the reused multi-mode resonator (MMR) with a coupled-line based absorptive network. The MMR is coupled to feeding lines to share its even-mode resonating points as transmission poles (TPs) in bandpass branch and transmission zeros (TZs) in bandstop branch. By employing the coupled-line based absorptive network, additional TZs and TPs are obtained in bandstop branch. Since these resonances in bandpass branch and the absorptive network are controlled independently, the bandwidth and centre frequency of the RTBPF can be tuned flexibly. The implemented RTBPF demonstrates a minimum insertion loss of 1.68 [email protected] GHz, 1.65 [email protected] GHz, 1.5 [email protected] GHz, three absorption peaks of − 6, −7 and −8 dB from dc up to 6 GHz. The overall circuit size is 0.22 λg ².

A rectifying surface for energy harvesting with optimal RF-to-DC efficiency insensitive to polarisation collimation is presented based on metasurface and branch-line coupler. The metasurface captures two orthogonal components of incident wave by two separate loads with high efficiency. A 90° branch-line coupler is introduced to redistribute two polarisation-dependent components into two invariant equipower outputs, aiming to preserve continuous impedance-matching between metasurface and rectifiers for maximised conversion efficiency. The proposed rectifying surface has been fabricated and measured. The measured fluctuation ranges of capture and conversion efficiency are smaller than 2 and 0.5%, respectively, agreeing well with authors’ prediction. They expect a wide range of applications to emerge from this novel concept in the future.

This Letter generalises frequency-domain oriented design guidelines for optimised single-resonant three-parameter AC current regulators for grid-connected converters. Unlike in previously established set of design guidelines, possible practical restrictions imposed by relatively low-switching frequency are taken into account in the proposed methodology. In addition, resonant term bandwidth is selected according to the desired robustness to fundamental frequency variations. As a result, a closed set of three non-linear equations is created. The solution maximises the loop gain crossover frequency for a given phase margin while respecting frequency robustness and resonant gain restrictions, set a priori.

In this Letter, a novel voltage regulation method is proposed for ensuring voltage security in photovoltaic (PV) distribution systems. It is a two-level regulation to reduce overall voltage deviation (VDE) and voltage difference (VDI). Firstly, evaluation indexes for VDE and VDI are built. Then, in order to reduce VDE, primary regulation by sensitivities-based power compensation is developed. Furthermore, in order to reduce VDI, secondary regulation by using PI coordinate compensator is built. Compared with other methods, the proposed method can achieve high regulation accuracy. The effectiveness is verified in a PV platform.

An auto-buck–boost single-inductor multiple-output (SIMO) dc–dc converter with duty-cycle-constrained comparator control is presented in this Letter. Combined both comparator-based and duty-cycle-constrained controllers enable the dc–dc converter attaining fast transient response and low cross-regulation (CR). The proposed algorithm is independent of the precision of current sensor and suppresses CR robustly with the proposed tri-feedback loop control of comparator-based voltage control, duty-cycle-constrained distribution and average current control. The proposed scheme is suitable for auto-buck–boost SIMO dc–dc converter design for dynamic voltage scaling applications.

A neural optical probe that is monolithically integrated with a micro-light-emitting diode (μLED), a micro-photodiode (μPD), and a cyclo-olefin polymer (COP) waveguide is a novel optogenetic tool that can suppress the influence of heat while utilising the advantage of LED. Although a μLED temperature increased by 0.3°C at a light output of 100 mW/mm², heat did not spread to the optical stimulation area. The emission from the μLED is propagated through a 2 mm long COP waveguide, suggesting that the μLED placed on the outside of the brain can optically stimulate the deep region of brain tissue. Furthermore, it is successfully detected at the μPD located close to the μLED, which proved that the μPD was effective to monitor the light output of the μLED.

This Letter deals with a practical signal processing problem where the filter is zero-phase and is specified by the frequency response. The authors show that if the signal is symmetrically extended, then the fast Fourier transform (FFT)-based algorithm can be efficiently implemented by using the discrete cosine transform without signal extension. They extend the proposed algorithm to 2D filters and show that computationally it is at least four times more efficient than the FFT-based algorithm. A practical application of the new algorithm is a fast Wiener inverse filter, which is used to estimate images taken under the condition of air turbulence.