A waveform generator for a 77 GHz automotive radar that uses multiple frequency shift keying (MFSK) modulation is presented. The use of this modulation can improve the detection ability of radar applications in multi-target situations. The proposed generator consists of a conventional 77 GHz Σ-Δ fractional-N synthesiser and MFSK modulation control logic (MMCL). The key innovation is the introduction of the MMCL, which has the ability to track the linear frequency increase while generating the frequency shift keying in each time step. Prior to realising the hardware implementation, a behavioural simulation was evaluated. In addition, a 77 GHz generator module was designed using 130 nm CMOS technology. In a chirp time of 6 ms, there were a total of 128 modulation steps. The bandwidth and frequency step of the chirp were 100 MHz and 300 kHz, respectively.

A dual-polarised, quadruple-ridged horn antenna with a coaxial feed operating over the frequency range of 75–500 MHz is presented. For the reduction of the antenna weight, a composite sandwich structure is used to build the antenna, which comprises solid aluminium and honeycomb materials. An adjustable shorting plate is put at the bottom of the antenna to prevent the occurrence of singularities created by standing waves. A Teflon plate is put at the centre position among the four symmetry ridges to mechanically support the huge antenna structure without breaking the symmetry of the ridges, leading to good electrical performance. Simulation and measured results are presented.

The design of a high-gain horizontally polarised antenna for ultra-wideband (UWB) operation is presented and experimentally demonstrated. Gain enhancement is achieved by using a frequency selective surface (FSS) reflector. Both the antenna and the FSS have simple structures and are printed on a similar substrate. The antenna is a three-stepped narrow rectangular slot excited using a three-stepped microstrip feed line. By optimising the lengths and widths of the sections of the slot and the feed line, an ultra-wide impedance bandwidth (3.2–12 GHz, measured) has been achieved. The FSS is a two-layer structure where each layer is an array of rectangular patches printed on one side of the substrate. With the application of the FSS, the antenna gain is enhanced by 3–4 dBi over most of the band. The polarisation purity (in terms of the measured cross-polar difference) is also improved with the FSS. Such an antenna–FSS combination will be useful for high-gain UWB applications requiring horizontal polarisation.

A high-power cavity-backed helical antenna with a unit ceramic radome is proposed and investigated for the first time. The helical antenna uses a unit ceramic radome to realise a vacuum seal for a single-element antenna, and the radome is designed in a thin-wall structure to maintain low insertion loss. The proposed antenna uses a backed cavity in a cylindrical structure to reduce the mutual coupling among elements. An S-band antenna prototype with a centre frequency of 2.856 GHz is designed and optimised. The performance of the antenna is simulated, including radiation performances and the power-handling capacity. This antenna is applied in an S-band 16-element sub-array and a 256-element array antenna. The experiment results validate that this antenna can realise favourable circularly polarised radiation in its axial direction in the frequency range of 2.7–3.0 GHz, and the average power-handling capacity for a single-element antenna is at least ∼4.7 MW.

A convex optimisation-based full polarimetric sum and difference patterns synthesis method for a conformal array is proposed. The method adopts a manifold separation technique and a co-polarisation and cross-polarisation basis to model the conformal array response in a closed-form expression from array calibration measurements, and then formulates the full polarisation sum and difference patterns synthesis as convex optimisation problems. The method is applied to a simulated conformal spherical conformal array composed of 176 antenna elements; the low sidelobe and low cross-polarisation level sum and difference patterns are presented.

A modified corner-fed dual-polarised stacked patch antenna, with improved port isolation, is proposed. By shifting the feed points along the lower patch edge and introducing a triangular cut in the middle of another edge, the isolation between the two ports is improved significantly. The proposed antenna achieves a return loss of about 14 dB and a port isolation of more than 28 dB, within the band 2.5–2.7 GHz. The proposed antenna is highly suitable for micro-base station applications in which features such as low profile, low cost and easy fabrication are required.

A newly designed two-dimensional planar inverted-F antenna (PIFA) that achieves a substantially enhanced dual-band impedance bandwidth is presented. It employs a proposed series-resonant feed structure, which consists of two series resonators. The enhanced dual-band impedance bandwidth is realised based on the appropriate increase of the impedance levels of the two series resonators. This increase is sufficient to cover wide use in mobile handsets of 698–960 MHz for the low band and 1710–2690 MHz for the high band with a VSWR < 3.

A dual-band, singly fed slotted microstrip antenna for the 2.4 GHz wireless local area network (WLAN) and 2.5/3.5 GHz worldwide interoperability for microwave access (WiMAX) is implemented. The antenna consists of an aperture-fed square patch embedded with a narrow and symmetrical π-shaped slot to generate dual-band resonance. The geometry resonates at a lower frequency band which fully covers the 2.4–2.484 GHz WLAN band and 2.5–2.69 GHz WiMAX band. The higher-frequency resonance is obtained in the 3.5 GHz WiMAX band. An oval-shaped stub at the open end of the single microstrip feed line is used to obtain impedance matching over a wide band of frequencies. The 10 dB return loss bandwidth measured for the antenna is 18.07 and 8.28% in the lower and upper bands, respectively. Reactive loading of the patch with the π-shaped slot has resulted in a reduction of patch size by 36% when compared with a traditional square patch resonating at an operating frequency of 2.4 GHz.

A scaled-down laboratory experiment has been performed in a well-controlled situation to find an enhanced rule for detecting a man-made underground tunnel by employing a cross-borehole frequency-swept radar system. A sleeve dipole antenna, fed by a coaxial cable passing inside its upper arm, is employed to easily put and pull it along an air-filled thin and long glass tube in scaled-down experiments. The tube-guided wave along the coaxial cable can be suppressed by ferrite-loading along the coaxial cable. However, the dipole antenna fed by the ferrite-loaded coaxial cable provides an asymmetrical radiation pattern in the near-field region. To improve the symmetry of its near-field radiation pattern, an additional ferrite-loaded wire is connected to the lower arm of the sleeve dipole antenna. According to the scattering measurement by an air-filled circular cylinder, the additional ferrite-loaded wire improves the symmetry of the electric field by more 30 dB in the frequency range of 1.5–3 GHz.

A molecular communication system operating in a pipe propagation channel with no induced flow is considered. Experimentally, it is shown that discrepancies in channel impulse response can be accurately modelled by an additive noise model. The noise amplitude is Nakagami distributed, and the shape and spread parameters of the distribution increase monotonically with propagation distance. Furthermore, demonstrated how the proposed noise model can be used to calculate the bit-error-rate and the capacity of a binary symmetric channel.

A novel integrated approach for modelling and control of a left ventricular assist device (LVAD) to achieve a physiological perfusion in patients with heart failure (HF) is proposed. A sensorless mathematical model of an LVAD has been developed to estimate an average pump flow. An advanced physiological control algorithm using the pole placement sliding mode control (PPSMC) approach was derived based on the estimated flow. The merits of the proposed approach have been evaluated with different physiological conditions using a cardiovascular model that is previously developed and validated using data from two healthy pigs. Simulation results demonstrate that the proposed control algorithm in conjunction with the mathematical model of pump flow is able to robustly track the desired reference pump flow with small error in the presence of disturbance and uncertainties.

The probabilistic switching of resistive random access memory (RRAM) can be utilised to implement physical unclonable functions (PUFs). By setting the operation condition at a switching probability of 50%, devices in a RRAM array are randomly settled into state ‘0’ or ‘1’ after programming. The RRAM switching probability provides a natural source of randomness that could be exploited in the PUF to generate security primitives. The feasibility and characteristics of the proposed PUF are analysed by simulation based on measured RRAM switching probability. With good scalability and stochastic mechanisms, RRAM may prove to be a promising candidate for security applications.

Transmission beyond 40 Gbit/s across a commercial Megtron 6, 11.5-inch electrical backplane is explored, and a record error-free serial electrical duobinary transmission of up to 48 Gbit/s including backplane connectors is demonstrated. Compared with non-return-to-zero (NRZ), the equalised duobinary transmission over the backplane demonstrates improved performance due to its higher bandwidth efficiency, and opens possibilities to increase the serial speed of backplane channels. Bit error rates are presented at different data rates and over different backplane channel lengths.

A second-order loop filter (LF) that uses a single op-amp resonator is presented for continuous-time delta–sigma modulators. The proposed technique improves the power and area efficiency of the LF and enhances the controllability of the transfer function and the resonating condition by providing a new RC network. A second-order modulator has been designed to confirm the effectiveness of the proposed scheme. Implemented in a 130 nm CMOS, the prototype modulator occupies an active area of 0.098 mm² and consumes 5.23 mW power from a 1.2 V supply. It achieves a peak SNDR of 60.95 dB over a 15 MHz signal bandwidth with a sampling frequency of 780 MHz.

A new non-rigid feature point matching algorithm is proposed, where the non-rigid global deformation is preserved by dividing the object into several overlapping local regions and only regularising the deformations of the local regions. A key property of the proposed algorithm is that the objective function is concave. By combining with the total unimodularity of the constraints, the proposed algorithms can be proved to directly yield integer valued solutions. Experimental results demonstrate the better robustness of the method over several state-of-the-art methods.

Presented is an architecture with mesh-structured electronic shutters and surround shutter drivers with positive/negative boosting to enhance the depth accuracy in three-dimensional (3D) depth sensors with a time-of-flight (TOF) measurement scheme. High-frequency modulation in the TOF scheme degrades the charge-transfer efficiency in high-resolution sensors because of the large parasitic resistance and the capacitance of the electronic shutters and shutter lines. This degradation suppresses modulation contrast, which induces depth errors. The global shutter-mesh architecture and surround-boosted driving scheme is implemented for high-modulation contrast in large-format depth sensors with a 1280 × 960 resolution.

A point target detection method based on clutter suppression is proposed to detect moving point targets in infrared (IR) image sequences. First, an algorithm with spatiotemporal local increment coding is proposed to estimate the clutter background. Then the target is detected by suppressing the estimated clutter background. During the spatiotemporal coding process, both inter-frame and intra-frame properties of the IR image sequence are taken into account. Experimental results demonstrate the validity of the proposed method.

A novel saliency model based on a discrete centre-surround (C-S) is proposed. By addressing the problem of filter scales selection commonly existing in the state-of-the-art saliency methods, the contrast map, obtained by a discrete C-S, can uniformly highlight salient objects. Furthermore, a saliency probability measure is derived from the combination of the discrete C-S and saliency priors, which further enhances the object's saliency. Experimental results on a dataset containing 1000 test images with ground truths demonstrate that the proposed model clearly outperforms previous saliency models.

A unified very large-scale integration (VLSI) architecture with butterflies that can perform photo core transform (PCT) in JPEG XR image compression is presented. The proposed architecture can achieve the unified architecture design, which supports the three elemental operations of PCT, and it has the characteristics of lower hardware cost, shorter critical path, lower power consumption, more efficient hardware utilisation and regular structure for VLSI implementation. Finally, the implementation on Altera field programmable gate array (FPGA) devices validates the effectiveness of the design.

A novel broadband chaos-based coherent binary phase shift keying (BPSK) communication system is proposed and it is demonstrated that its theoretical probability of error expression matches that of a conventional narrowband BPSK system. The result is confirmed with the corresponding empirical bit error rate (BER) simulation. Most importantly, it is shown that the proposed system outperforms the chaotic communication systems in terms of BER while also being more practicable. The proposed system thus holds high promise as the chaotic communication systems pose implementation difficulties, lack robust synchronisation techniques and are generally considered of academic interest only.

A novel half-mode substrate-integrated waveguide (HMSIW) bandstop filter loaded with a half complementary split-ring resonator (HCSRR) is investigated. The experiment indicates that the novel HCSRR shows almost the same resonant characteristic but half size in contrast with the original CSRR. Then the proposed HCSRR is etched on the middle of the HMSIW and a compact bandstop filter is realised. A practical example for a HMSIW bandstop filter is fabricated and measured. The measured results show good agreement with the simulated ones.

Broadband line drivers usually include a back termination resistor at the output to match to the cable they are driving. Half the output power is dissipated in this resistor, therefore halving their efficiency. The resistive termination can be replaced with synthetic impedance using a combination of voltage and current feedback. This implementation is based around a current mirror amplifier capable of producing a 40 V peak-to-peak signal into a 50 Ω load impedance from a ±24 V supply. Using bandwidth enhancement, the small signal bandwidth was 15 MHz and the full power bandwidth was 6 MHz. When driving a 50 m long terminated 50 Ω coaxial cable, the amplifier produced a clean output signal with little ringing, indicating that it is well matched over a broad bandwidth.

A compact and high frequency selectivity two-stage microstrip bandpass filter is designed and experimentally demonstrated. Such a kind of filter is composed of two pairs of twist-modified asymmetric split-ring resonators (SRRs) to achieve three transmission poles within the passband and four transmission zeros at low and high rejection bands. Sharply out-of-band reductions contributing to the high frequency selectivity are obtained by properly detuning the gaps among the SRRs. The measured results show that the proposed filter with dimensions of 0.088 × 0.088λ ² has a 3 dB passband of 15.2% centred at 1.98 GHz. This filter can be flexibly integrated in miniaturised radio-frequency/microwave circuits.

A new balun–bandpass filter with a quasi-elliptic response for 60 GHz wireless communication systems is proposed. It is designed in the form of a substrate integrated waveguide (SIW) and fabricated with the standard low-temperature co-fired ceramic (LTCC) technology. Both simulated and experimental results are provided, which have good agreement. The results indicate that the proposed balun filter has a good balance performance with <1.1 dB amplitude imbalance and better than 3.4° phase imbalance, compact size, as well as high selectivity.

A simple wideband tubular suspended stripline (SSL) bandpass filter using novel miniaturisation technology is presented. In this design, the equivalent circuit of the proposed filter is considered as the conventional tubular bandpass filter circuit. A new miniaturisation technology called ‘different height of cavity’ is applied to the filter design. A SSL filter is designed and fabricated with the passband from 0.6 to 1 GHz. The insertion loss is better than 0.6 dB in the passband while the return loss is mostly higher than 20 dB. The group delay variation is <0.2 ns. Good agreement between the simulated and measured result is observed, which proves the validity of the technique.

The first results of an integrated hybrid photo-receiver based on a 850 nm two-terminal (2T) silicon germanium (SiGe) heterojunction bipolar photo-transistor (HPT) for low-cost radio-over-fibre (RoF) applications is presented. A hybrid module was realised with two cascaded low-noise amplifiers of 20 dB total gain and an SiGe 2T-HPT exhibiting a −15 dB opto-microwave gain at 5.15 GHz with a 1.6 GHz bandwidth. A 16% degradation of the error vector magnitude compared with back-to-back measurement for the transmission of a 2 GHz OFDM signal at 3 Gbit/s was measured based on the 60 GHz IEEE 802.15.3c standard transposed at the intermediate frequency (IF) of 5 GHz, according to an IF-RoF distribution mode.

A new calibration method for dual-mask liquid crystal spatial light modulators (SLMs) is presented, which can be applied while the SLM is in the setup of a pulse shaper optimised for semiconductor laser sources. An evolutionary algorithm is used to retrieve the phase retardances from the transmitted spectral intensity.

Pentacene organic thin-film transistor (OTFT) with HfYO as gate dielectric has been fabricated at room temperature on adhesive flexible vacuum tape and shows high performance. The threshold voltage is as low as −1.77 V due to the high-κ gate dielectric employed. A small sub-threshold swing (SS) of 0.145 V/dec indicates good interface between the gate dielectric and pentacene film. Atomic force microscopy (AFM) reveals that large pentacene gains form on the HfYO gate dielectric, leading to a high carrier mobility of 0.236 cm²V⁻¹s⁻¹. These good results suggest that the adhesive vacuum tape can be used as the substrate for adhesive flexible OTFT, and HfYO is a promising gate dielectric for high-performance pentacene OTFT.

Attributed scattering centres in synthetic aperture radar imagery reveal information that would otherwise be unknown to an analyst. How the method used for pixel qualification affects the attribution is considered. Using the well-known Prony method as the standard, the effects of position estimation from local peak detection and the two-dimensional (2D) Prony method on pixel attribution to canonical scattering centres are shown. The efficacy of using 3D attributed scattering centres to characterise a commercial vehicle is also demonstrated.

0.25 μm emitter InP/InGaAs double heterojunction bipolar transistors (DHBTs) are presented that simultaneously exhibit a high collector current density (J _c) of more than 3 mA/μm² and high breakdown voltage (BV_CEO) of 12 V. The DHBTs consist of a 30 nm-thick InGaAs base, 250 nm-thick InGaAs/InAlGaAs/InP collector and 150 nm-thick n-doped InP field buffer. Since the doping level of the InP field buffer is relatively high, only the InGaAs/InAlGaAs/InP collector is depleted at low V _CE. Thus, the DHBTs can provide f _t = 173 GHz and f _max = 470 GHz at J _c = 3.5 mA/μm². On the other hand, at a high V _CE, both the InGaAs/InAlGaAs/InP collector and InP field buffer are depleted. Therefore, the effective depletion thickness increases, which results in a BV_CEO of 12 V. These results indicate that the use of the InP field buffer provides both high-speed performance and high BV_CEO.

Close to glottal closure instants (GCIs), the speech signal is expected to change its amplitude rapidly and, at GCIs, it is expected to have strong negative peaks. A novel algorithm that exploits these two properties for the estimation of GCIs is presented. Here, a symmetrised speech segment is assumed to be a Fourier transform (FT) of an even function. In such a case, at the locations of the GCIs, the strong negative peaks in the symmetrised speech segment correspond to zeros that lie considerably outside the unit circle in the z-plane. The group delay spectrum of the time-domain signal derived by taking inverse FT of this assumed FT is expected to take a value close to −2π at the angular locations of these zeros. Mapping frequency scale to time scale, the frequency bins for which group delay reaches −2π correspond to the locations of GCIs. Theoretical justification for the proposed approach is also presented by defining a novel function called the conditional group delay function. Systematic evaluation is carried out on the CMU Arctic database and the performance of the proposed technique is better than that of the algorithms namely DYPSA, ZFF, YAGA and is close to that of SEDREAMS.

Sparse deep neural networks (SDNNs) for speaker segmentation are proposed. First, the SDNNs are trained using the side information that is the class label of the input. Then, speaker-specific features are extracted from the super-vector feature of the speech signal by the SDNNs. Lastly, the label of each speech frame is obtained by K-means clustering, which is used to segment different speakers of a continuous speech stream. The performance evaluation using the multi-speaker speech stream corpus generated from the TIMIT database shows that the proposed speaker segmentation algorithm outperforms the Bayesian information criterion method and the deep auto-encoder networks method.

A self-tuned surface acoustic wave (SAW)-less global navigation satellite system (GNSS) receiver is proposed. In the self-tuning mode, the radio-frequency amplifier (RFA) is configured as an oscillator and the receiver chain is reused to realise a simple self-calibration. In the normal operation, the RFA works as a Q-enhanced radio-frequency filter to remove the out-of-band blockers. A gain-irrelevant DC offset cancellation is also presented. This GNSS receiver achieves 4.8 dB noise figure and a rejection of 25.8 and 31.8 dB for DCS (1710 MHz) and PCS (1850 MHz) blockers, while consuming <5 mA current from a 1.7 V supply.