A novel compact printed antenna with a spiral winding and a loop feed is proposed. After optimising the design parameters, the prototype of the antenna with a small electrical size of 0.043λ_c × 0.043λ_c × 0.001λ_c is fabricated. It is shown that a central frequency as low as 425.9 MHz can be achieved without any additional matching networks, and a good omnidirectional radiation pattern similar to the dipole antenna is obtained.

A frequency-scanning grating-reflector antenna operating in the 0.2 THz band is proposed. The antenna system consists of a planar grating reflector, a thin lens and a feed horn placed at the focal point of the lens. The beam-scanning property based on the frequency sweeping was analysed and optimised based on Floquet's theorem. A bilayer structure is introduced to improve the diffraction efficiency and to suppress the specular beam. The frequency-controlled beam steering concept was verified by the quasi-optical measurements in the 0.2 THz band. The measured scanning range is 13.55° over the frequency band 180–220 GHz. The specular beams are remarkably inhibited with the diffractive efficiency as high as over 89%. The presented antenna has the advantage of the convenience to be extended to a higher frequency with a high directivity.

A small electrical metamaterial (MTM) antenna, based on a modified electrical coupled resonator for bandwidth enhancement, is presented. The bandwidth enhancement is achieved using a modified electric coupled (MELC) resonator as the antenna's main radiating element. This modification is achieved using two strip lines placed below a conventional electric coupled (CELC) resonator. The measured impedance bandwidth at −10 dB is 20.7% at a resonance frequency of 2.62 GHz. The design is validated by experimental results, and the proposed antenna is suitable for worldwide interoperability for microwave access (WiMax) applications (2.5–2.69 GHz).

A new ultra-wideband (UWB) antenna integrated with a narrow-band (NB) antenna has been designed and fabricated. The UWB antenna consists of a U-shaped dielectric resonator fed by a microstrip line. In addition, the proposed structure integrates an NB slot with the same polarisation. The electromagnetic analysis is carried out by using two commercial software tools. To validate the proposed concept, experimental measurements have also been performed. The measured results show that the dielectric resonator antenna (DRA) provides a wide bandwidth from 2.94 to 11.34 GHz (covering the UWB spectrum, 3.1–10.6 GHz) while the slot antenna provides a 2:1 voltage standing wave ratio bandwidth for 5.71–6.13 GHz. Moreover, the proposed design allows an efficient integration with very good isolation between the two ports (transmission coefficient S ₂₁ <−20 dB in the whole operating frequency band).

It is shown that memcapacitive (memory capacitive) systems can be used as synapses in artificial neural networks. As an example of the proposed approach, the architecture of an integrate-and-fire neural network based on memcapacitive synapses is discussed. Moreover, it has been demonstrated that the spike-timing-dependent plasticity can be simply realised with some of these devices. Memcapacitive synapses are a low-energy alternative to memristive synapses for neuromorphic computation.

A thin, flexible and skin-attachable sensor for continuous monitoring of blood pressure is reported. The sensor was fabricated as a layered structure of ferroelectret film, specially designed electrodes and a flexible electronic circuit, which together enable simultaneous measurement of ballistocardiogram and an electrocardiogram on the human chest with minimal discomfort. To investigate the feasibility of continuous blood pressure monitoring, systolic blood pressure was estimated using the RJ interval, which is computed as the time delay between the R peak of the electrocardiogram and the J peak of the ballistocardiogram. The results indicate that the estimated systolic blood pressures are in excellent agreement with the reference values, achieving an intraclass correlation coefficient of 0.95 (P < 0.01), and complying with the recommended guidelines from the Association for the Advancement of Medical Instrumentation in the United States and the British Hypertension Society.

Implantable and portable medical devices have attracted much interest in recent years and the interconnections of these devices have become an important issue. One interconnection method is the use of the body access network (BAN) that does not need high power consumption as conventional radio frequency communication methods. Previous studies of the BAN have focused on measuring between the chest, wrist and other surface parts of the skin which are easily accessible with conventional electrodes. Deformable BAN electrodes are implemented to study the effectiveness of BAN communications in the ear canal. Conventional earphones and Bluetooth devices can easily access the ear canal and could be used for the BAN communication channel. The experimental results show various transmission loss among the ear canal, the wrist and the chest. The lowest attenuation showed as 23 dB at 34 MHz.

A more rapid detection method for step initiation is an important challenge for muscle force reduction in lower limb exoskeletons. Many studies have attempted to detect step initiation faster via either the precedence walking assistance mechanism or the shadow walking assistance mechanism. Although the precedence walking assistance mechanism is the fastest for the detection of step initiation, this mechanism has limitations while using biosignals. As a result, many studies have attempted to detect step initiation with the shadow walking assistance mechanism as soon as possible. A precedence walking assistance mechanism has been implemented for step initiation in the Unmanned Technology Research Centre Exoskeleton (UTRCEXO) based on gait analysis without using any biosignals. Success was achieved in reducing the muscle force of the operator.

An optimal pulse compression scheme is proposed for the synchronised time-lens source for coherent Raman scattering (CRS) imaging, using CRS signals as the criterion. The simulation results show that instead of compressing the pulse width of the time-lens source to its minimum, peak power should be optimised to maximise CRS signals. In terms of experiment, this can be done by optimising the second-harmonic signal or two-photon current of the time-lens output.

A linearised differential common-gate CMOS low-noise amplifier (LNA) is proposed. The linearity is improved by a cross-coupled post-distortion technique, employing PMOS in a weak inversion region as an auxiliary field effect transistor to cancel the third-order nonlinear currents of a common-gate LNA and impair the related second-order nonlinear currents. Meanwhile, the resulting noise figure is little affected. The LNA implemented in a 0.18 μm CMOS technology demonstrates that IIP3 and gain have about 8.2 and 1.5 dB improvements in the designed frequency band, respectively. A NF of 3.4 dB is obtained with a power dissipation of 6.8 mW under a 1.8 V power supply.

A dual-phase pulse-modulated polar transmitter (PMPT) for wide power control range (PCR) applications is presented. For the validation of the proposed architecture, a radio-frequency (RF) PMPT chip is implemented with a standard 0.18 μm 1P6M mixed-signal CMOS process. The measurement results show that by using the proposed technique, wide PCR transmitters can be implemented accurately with low hardware requirements. An RF variable-gain amplifier with just 21 dB of dynamic range is sufficient to pass the 80 dB PCR of the CDMA2000 standard. Compared with a conventional transmitter utilising linear power amplifiers, the proposed architecture has the potential of achieving much higher efficiency while being able to achieve chip-level integration with eased circuit design specification.

A four-scale continuous wavelet transform filter bank implemented in a 0.18 μm CMOS process is presented. Experimental results demonstrate the real-time multi-scale analysis of electrocardiogram signals with a power consumption of 1.3 nW, the lowest reported power consumption for a complete wavelet filter bank. In addition, the results demonstrate for the first time the use of pico-amp scale currents in a 0.18 μm CMOS process for providing advanced signal processing functions.

An improved reference voltage generation scheme is proposed for a 1T1C-type ferroelectric random access memory (FRAM), in which the circuit referring to reference cells is redefined and the data are written into reference cells at random between ‘1’ and ‘0’ depending on the voltages of the bitlines during every operation cycle. Compared with conventional schemes, it can not only realise higher access speed for memory, but also can enhance its reliability by resolving the imprint and relieving the fatigue relating to ferroelectric capacitors in the device. Functional verification for the experimental prototype utilising the proposed scheme has been implemented.

A new single-bit ΔΣ modulator based on multi-bit quantisation and a successive approximation (SA) algorithm with improved output swing due to a reduced quantisation noise power is presented. Although this structure maintains the linearity of the single-bit digital-to-analogue converter (DAC), it also removes the requirement for a high slew rate operational amplifier. Hence, the modulator dissipates less power.

For processing charge coupled device (CCD) signals, a correlated double sampling (CDS) circuit must be used to eliminate noise. Usually, the amplitude of the CCD signals is uncertain, either too small or too large, hence a variable-gain amplifier (VGA) is also needed to amplify the CCD signals. Traditionally, the CDS and the VGA are separated, each of which consumes much energy and large area. If only the CDS and the VGA would be combined together, much energy and area could be saved. Hence, a CDS with VGA functionality is proposed to achieve this goal. Besides, the VGA functionality provides a gain range of 0–18 dB, programmable with a 9 bit resolution, which is far better than the traditional ones. Also, the VGA gain curve follows a linear-in-dB characteristic, hence the VGA functionality has rather good linearity. Finally, the VGA gain curve demonstrates the good performance of the VGA functionality.

A new compensation method for low dropout (LDO) voltage regulators is proposed, where the series resistor of the conventional Miller compensation changes with the load current to track the variations in the first non-dominant pole.

A novel class of multi-stage noise-shaping ΣΔ modulators is presented. The proposed structure achieves a one-order higher noise-shaping than the modulator's order only by using two extra analogue paths. The noise transfer function zero optimisation is achieved by adjusting a single coefficient in the proposed modulator. Hence, with a minimal extra circuit, the signal-to-noise and distortion ratio is significantly increased compared with the conventional cascaded modulator. Analytical calculations and simulation results are provided to verify the usefulness of the proposed modulator.

A new programmable delay-locked loop (DLL) based fractional frequency multiplying clock generator is presented. In contrast to conventional DLL-based clock generators that generate only integer clock multiplication, the proposed clock generator provides fractional-ratio frequency multiplication while maintaining the advantages of DLLs, such as the deskewing between the input and the output clocks. Implemented in a 0.13 µm 1.2 V CMOS process, the proposed clock generator achieves an effective peak-to-peak jitter of 7.5 ps and occupies an active area of 0.018 mm² while dissipating 9.0 mW at 1.5 GHz. The output frequency ranges from 0.85 to 1.5 GHz with programmable fractional multiplication ratios of N/M, where N = 4, 5, 8, 10 and M = 1, 2, 3.

Tiny is a scalable and efficient three-dimensional (3D) network-on-chip (NoC) designed to reduce latency and area. A theoretical analysis demonstrates its efficiency when compared with a basic 3D mesh NoC. Mapping independent traffics with different injection rates makes the trade-offs analysis of Tiny possible. Results highlight that Tiny always reduces area and for several cases minimises latency.

A new low-phase-noise colpitts LC quadrature voltage controlled oscillator (QVCO) is presented. The proposed QVCO is realised by coupling two differential current-switching Colpitts oscillators in an ‘in-phase–anti-phase’ manner. The coupling technique between the two Colpitts LC-voltage controlled oscillators (VCOs) has significantly improved the phase noise of the proposed circuit without using any extra elements. The oscillation frequency of the proposed LC-QVCO is 4 GHz. The Colpitts QVCO exhibits a phase noise of −131.5 and −141.3 dBc/Hz at 1 and 3 MHz offset, respectively. By generalising the coupling scheme to more than two differential Colpitts VCOs in a loop, multiphase signals can be generated.

A series electrostatic discharge (ESD) diode structure is proposed to minimise the degradation induced by thermal interaction between series connected diodes. The proposed series diode is constructed using a distributed cell-based ESD diode. To verify the feasibility of the proposed structure, single and series diodes are designed using the typical and the proposed structures. From the experimental results, it is proved that the ESD survival levels of the proposed series diode are nearly identical to those of the single diode, unlike the case of typical series diodes.

A depth map upsampling method using depth local features is proposed. Depth discontinuity and depth variance are extracted from a low-resolution depth map and a colour image as the depth local features. They are incorporated into an energy function for the Markov random field (MRF)-based depth upsampling. The high-resolution depth map is obtained by optimising the energy function using belief propagation. The experimental results show that the proposed method outperforms other depth upsampling approaches in terms of the bad pixel rate.

A goodness compactness measure is necessary in computer medical diagnosis, shape analysis and computer vision processes. A compactness measure called the normalised E-factor (NEF) is introduced. It is shown that this measure is simple and robust to translations, rotations and scale changes and that it satisfies the set of criteria for a good compactness measure. Through a series of experiments it is shown that the normalised E-factor is useful for shape description, measuring digital compactness with or without holes and that it overcomes some drawbacks that are present in the classical and normalised discrete compactness measures.

A new method of obtaining geometric features of three-dimensional range data for human and tree recognition in an off-road environment is described. The learning algorithm AdaBoost is used to select a set of discriminative features from a very large set of potential geometric features. The proposed geometric feature can be considered as a generalisation of the geometric feature used in previous studies. The experimental results for human and tree recognition show that the proposed method outperforms the other methods.

Fingerprint enhancement, which aims to improve the clarity of fingerprint ridge structures, plays a very important role in automatic fingerprint recognition. Recently, a novel enhancement method called curved Gabor was proposed, which achieves a state-of-the-art enhancement effect, but its high computation becomes a bottleneck. An orthogonal curved-line Gabor filter is proposed to overcome this issue. This method decomposes the unrotated two-dimensional (2D) Gabor into a 1D bandpass Gabor and a 1D lowpass Gabor, then implements them with two separable steps: a 1D bandpass Gabor and a 1D lowpass Gabor filtering on curved lines, respectively, perpendicular and parallel to the ridge orientation. The experiments show that the method is about 42 times faster than the curved Gabor while retaining comparable enhancement effect.

A subband adaptive filtering algorithm is proposed which improves its performance by adjusting step sizes. The proposed algorithm derives the individual step sizes for each subband instead of using a common step size for multiple subbands. The derivation of the step sizes is based on the mean-square deviation minimisation in order to achieve the fastest convergence at the instant. Furthermore, the individual step sizes contain the squared norm of the input vector, hence it leads to the regularisation effect that helps the algorithm work well in the case of badly excited input signals. The simulation results show that the proposed algorithm achieves a faster convergence rate and a smaller steady-state estimation error than the existing algorithms.

A novel data-nulling superimposed pilot scheme for orthogonal frequency division multiplexing (OFDM) systems is proposed, where the input data vector is spread over all the subcarriers by a precoding matrix and then nulled at certain subcarriers for the insertion of training pilots. This method avoids the loss of the data rate for frequency-division multiplexed pilots, but results in the distortion of input data. To mitigate the distortion introduced by the nulling operation, a simple iterative reconstruction scheme is used to improve the detection performance.

A new high-order differential space-time block code (DSTBC) is introduced as an attractive alternative to the well-known coherent space-time block code (STBC) by providing both full space-time diversity and blind detection. A differential detection based on the orthogonal STBC scheme is proposed. The proposed differential scheme is a generalisation of the recent DSTBC already designed for two transmit antennas. The proposed differential encoder consists of a specific trellis diagram that creates a positive correlation between successive differential data. At the receiver side, the well-known Viterbi algorithm is applied using a specific metric that nearly compensates for the 3 dB penalty commonly introduced by the symbol by symbol differential schema, and shows a 0.5 dB performance loss only from the coherent STBC scheme.

Rendezvous is the process in which two users attempt to meet on a common channel and it is of fundamental importance in cognitive ad hoc radio networks. A cognitive radio blind rendezvous algorithm named simple role-based is proposed that exploits the common channel labelling inherent to wireless networks. The scheme guarantees the simultaneous rendezvous of multiple cognitive users. Analytical expressions are derived for the expected time to rendezvous and maximum time to rendezvous. Numerical results show that the proposed method outperforms the other well-known rendezvous algorithms.

A 1.7/2.6 GHz high-efficiency concurrent dual-band power amplifier (DBPA) with a dual-band harmonic wave controlled transformer is demonstrated. The proposed dual-band transformer, which transfers the fundamental and the second harmonic impedances of each frequency, is directly transmission line implemented with the help of an optimisation algorithm. The measurement results show that the proposed DBPA has performances of 70 and 64% power added efficiency with an 11 dB gain and a 41 dBm output power at 1.7 and 2.6 GHz, respectively. Its average drain efficiency is more than 42% and the adjacent channel leakage ratios are lower than −46 dBc at both frequencies with the digital predistortion adjustment for a 5 MHz WCDMA signal input.

A complementary metal oxide semiconductor (CMOS) downconversion mixer for ultra-wideband (UWB) applications is presented. The mixer circuit is designed using a VIS 0.25 μm radio frequency (RF) CMOS technology, working at the 0.9–10.6 GHz frequency range; it will be used in applications such as IEEE wireless fidelity (Wi-Fi), global systems for mobile (GSM) and worldwide interoperability for microwave access (WiMAX). The core of the mixer has been designed based on double-balanced cell architecture, and uses the current-bleeding method to increase the linearity and improve the conversion gain. A resistor is put on the drain of the MOS between two RF inputs, which will improve the flatness of the conversion gain. The measured conversion gain of the mixer is 8.6 ± 1 dB. The 1 dB compression point of the intermediate frequency (IF) output power is equal to −8 dBm. The RF and the local oscillator (LO) input return loss are well below −9 dB. The noise figure is 13.1 dB, whereas the IF is 100 MHz, and the mixer core dissipates 9.8 mW under a 1.8 V supply. The analytical, simulated and measured results are mutually consistent.

A novel dual-band bandpass filter with a compact structure and a wide stopband is proposed. The total structure includes coupled-line stubs and coupled-line connected sections, leading to a compact circuit. This proposed filter can be implemented easily by using single-layer microstrip technology. The measured results of a prototype filter show that it can operate at both 1.32 and 2.67 GHz and up to the 8 GHz 15 dB stopband.

A composite frequency-selective surface (FSS) fabricated by using an e-beam evaporator is proposed. This manufacturing process is advantageous in that it can directly achieve an FSS function for real structural surfaces with various skin conformations. A multilayered patch–grid–patch-type FSS consisting of copper layers as capacitive or inductive surfaces and a glass/epoxy prepreg as the substrate is produced. It operates at 9–10 GHz for dual polarisation. This technique can be used to manufacture a stealth technology for communicating in a specific frequency range.

Phase delay characteristics of the complementary triangular split ring resonator (CTSRR) loaded transmission line with capacitive stubs are presented. The phase responses of a reference microstrip transmission line and a transmission line with capacitive stubs, loaded with the CTSRR on the ground plane, are compared. The practical results show that the CTSRR loaded transmission line provides four times more phase delay when compared with a reference transmission line, and thus the proposed configuration provides a 76% transmission line length reduction.

A fourth-order fully canonical dual-band bandpass filter (DB-BPF) using λ/4 stepped impedance resonators (SIRs) is presented. In the mainline path, four dual-band λ/4 SIRs are serially coupled by alternative J and K inverters, and cross-couplings, K ₁₄ and J _SL, are analytically introduced to achieve two pairs of finite transmission zeros at both sides of each passband. To experimentally demonstrate this attractive fully canonical dual-passband filtering performance, a fourth-order microstrip DB-BPF centring at 1.8 and 5.8 GHz with two pairs of finite transmission zeros for each passband is synthesised and fabricated.

A wideband probe-type microstrip-to-waveguide transition integrated in a low-temperature co-fired ceramic (LTCC) is presented and experimentally verified. To enhance the bandwidth, an air-hole matching structure besides the probe is buried in an LTCC substrate, which is implemented by drilling square holes on different substrate layers. The microstrip-to-waveguide transition is fabricated to verify a design with compact size. The measured results show a good agreement with simulated results.

A novel superchannel transmission scheme is proposed where an optical comb source is used to carry the multiband signal and one subcarrier is reserved for regeneration as a local oscillator at the receiver end. By combining carrier tracking and carrier expanding, a 4 × 2 Tbit/s superchannel wavelength division multiplexed self-coherent transmission over a 320 km standard single mode fibre has been successfully demonstrated.

A long-standing goal of GaN device research has been the development of a reliable, well-controlled process for p-GaN formation by ion implantation. Results to date have indicated an activation of 1% or less using high-temperature rapid thermal annealing (RTA) techniques and coimplantation. Although Mg is a relatively deep acceptor, this is still much less than the theoretically achievable value (8.2% based on the 160 meV acceptor level). A multicycle RTA process is presented that is capable of achieving up to 8% activation of the Mg-implanted GaN. This approaches the theoretical value, and represents a significant step in GaN device research.

A numerical model for parametric amplifiers, which include stochastic variations of the group velocity dispersion (GVD), is presented. The impact on the gain is investigated, both with respect to the magnitude of the variations and by the effect caused by changing the wavelength of the pump. It is demonstrated that the described model is able to predict the experimental results and thereby provide a quantitative evaluation of the standard deviation of the GVD. For the investigated fibre, a standard deviation of 0.01 ps/(nm km) was found.

A simple scheme is proposed to increase the hold voltage and not change the original trigger voltage of a silicon-controlled rectifier (SCR) to enhance its latch-up immunity without changing the device dimensions. It is found that using the lightly doped P-diffusion instead of the highly doped P⁺ diffusion as the P emitter of the anode can increase the hold voltage of an SCR.

What limits performance in some recently published digital floating level-shifters are identified and a new level-shifter is proposed to address these limitations. An order-of-magnitude reduction in power dissipation compared with the recently published fast high-voltage level-shifters is obtained; simulations in a high-voltage 0.35 μm process show the proposed design is capable of shifting a 2.5 V logic signal up by 17.5 V with a propagation delay of 3 ns and a transition energy of 6 pJ.

The synthetic aperture radar (SAR) imaging methods based on compressive sensing endure great computational complexity since they have to deal with a large measurement matrix when processing the raw data. An efficient ℓ _q regularisation algorithm for SAR imaging is proposed, which uses a decoupling operator constructed by means of chirp scaling. It can reduce the computational complexity from the square order to the linear logarithm order compared with the conventional ℓ _q regularisation-based imaging algorithm while preserving the equivalent imaging quality. The proposed algorithm is efficient in the sense that it allows reconstruction of the sparse scenes with the sub-sampling data.

A bistatic chirp scaling algorithm (CSA) is presented to process the forward-looking bistatic synthetic aperture radar (FL-BiSAR) data. First, the double-square-root equation in the general bistatic range model can be simplified to a square-root equation, when the size of the scene is small enough. Based on the model, the range Doppler spectrum is derived. Then, a CSA is obtained. The range model is proved to be an effective proxy by a simulation.

A low-power and wideband three-stage 77–81 GHz low-noise amplifier (LNA) using standard 90 nm CMOS technology is reported. The T-match technique is utilised to achieve simultaneously wideband input and output impedance matching, wideband power gain (S ₂₁) and a wideband noise figure (NF). The LNA consumes 21.1 mW, achieving S ₁₁ better than −10 dB for the frequencies 62.3–82.4 GHz, S ₂₂ better than −10 dB for the frequencies 62.8–84.6 GHz, S ₁₂ better than −29 dB for the frequencies 72–84 GHz and a group delay variation smaller than ±4 ps for the frequencies 77–81 GHz. In addition, high and flat S ₂₁ of 13.1 ± 1.5 dB is achieved for the frequencies 72–84 GHz, which means the corresponding 3 dB bandwidth is 12 GHz. Furthermore, the LNA achieves a minimum NF of 6.2 dB at 78 GHz and a NF of 6.8 ± 0.6 dB for the frequencies 75–82 GHz, one of the best NF results ever reported for a CMOS LNA for 77–81 GHz automotive radar applications.

An ultra-low R _on,sp 700 V DB-nLDMOS (dual P-buried-layer nLDMOS) which uses 0.35 μm technology and full ion implantation technology is proposed. Experimental results show that with 800 V BV _ds, R _on,sp is only 10.7 Ω · mm² which is the lowest value of triple RESURF (REduce SURface Field) LDMOS reported before. This mainly benefits from two aspects. First, thermal budgets of the process are strictly limited after implantation of the Pbury layer. Secondly, device sizes of the neck region are optimised to reduce R _on,sp which also suppress the JFET effect of the triple RESURF LDMOS.

An AlN/AlGaN high-electron-mobility transistor (HEMT) fabricated on a free-standing AlN substrate is demonstrated. A metal stack, composed of Zr/Al/Mo/Au, was found to show low contact resistivity for source and drain ohmic contacts. The fabricated AlN/AlGaN HEMT exhibited a maximum drain current of 38 mA/mm with a threshold voltage of −3.4 V. Negligible drain current degradation was observed at temperatures from 300 to 573 K, demonstrating that an AlN/AlGaN approach on an AlN substrate is promising for stable high-temperature operation.

The bipartite matching strategy is first applied in the graph theory to optimise the resource allocation in a device-to-device (D2D) underlaying cellular network. With each D2D group reusing the resource of one particular cellular user, the proposed scheme minimises the cross-interference between D2D users and traditional cellular users with polynomial complexity. Simulation shows that the optimal system capacity boundary is achieved by the proposed scheme and D2D communication shows its great potential in increasing system capacity.

A new frequency-hopping sequence (FHS) set is presented and studied. The construction of new FHSs is based on the polynomial theory over the finite field, and the new FHSs are called the power residue and the constant FHSs. It is shown that the presented FHS set is optimal with respect to the average Hamming correlation bound and it includes the known cubic and constant FHS set as a special case.

A charge-domain downconversion sampling mixer with an embedded four-path bandpass filter for band selection and a sinc³ finite impulse response (FIR) and a first-order sinc FIR with a decimation of 4 for channel selection is proposed to provide sufficient attenuation of out-of-band blockers and aliasing interferences. The sampling mixer designed in an IBM 130 nm 1.2 V CMOS with a 100 MHz input exhibits an 80 dB suppression of aliasing interferences and a 57 dB stopband attenuation while consuming 1.08 mA.

A new type of tiered code for designing the ranging signal structure of global navigation satellite systems (GNSSs) is proposed. The proposed tiered code structure combines a constant amplitude zero autocorrelation polyphase secondary code with its primary code, resulting in a tiered polyphase code (TPC) that avoids cross-correlation interference between different ranging codes. The analysis and the simulation results show the performance of the proposed TPC and confirm its effectiveness as a ranging signal structure for the GNSSs.

The design and implementation of dual-band duplexers is presented. The proposed method works well for duplexers with narrow bandwidths and it gives the designers the flexibility to achieve desired return loss levels at different frequency bands. To verify the design concept, dual-band duplexers based on cavity-type resonator topologies are designed, fabricated and tested. Good agreement is achieved between the theoretical synthesis results, the simulation results and the measurement results of the duplexer prototype.

The impact of a link adaptation feedback interval on user net throughput for time division multiple access (TDMA)-based single-carrier very high frequency (VHF) narrowband mobile ad-hoc networks (MANETs) is investigated. Specifically for the narrowband MANETs optimisation of the signalling overhead is one of the major challenges. Reducing the overhead means that more user data can be transmitted; however, the quality of channel quality feedback information will be lower, resulting in potentially inaccurate link adaptation decisions. The results from Monte Carlo computer simulation show that a global maximum for the throughput with respect to interval exists. The achievable throughput, however, strongly depends on the user mobility pattern.

Optimum antenna combining is utilised to cope with the transmitter (TX) leakage issue encountered in carrier aggregation. By assuming the transmission of pilot symbols and exploiting the structural information of the noise-plus-interference covariance, the maximum likelihood covariance estimation is mathematically derived. The simulation results verify that the proposed antenna combining with the structured covariance estimation provides a noticeable performance gain in the presence of TX leakage.

It is shown that by exploiting the degrees of freedom arising from the use of multiple antennas and/or insertion of virtual carriers in multiple-input multiple-output (MIMO) orthogonal frequency-division multiplexing (OFDM) systems, perfect symbol recovery and complete narrowband interference (NBI) cancellation can be achieved in the absence of noise by means of constrained minimum mean-output-energy linear filtering. Moreover, a nonlinear iterative scheme is proposed to improve NBI rejection in realistic operative conditions.

The performance of channel adaptive random access is investigated with imperfect channel reciprocity between the forward channel gain at the measurement time and the reverse channel gain at the random access packet transmission time.

For simultaneous wireless information and power transfer (SWIPT), secure transmission is an important issue. The power allocation problem is investigated to maximise the achievable secrecy rate in a downlink orthogonal frequency division multiple access system with a power splitting SWIPT scheme. Unlike the traditional wireless communication systems, it is assumed that the legitimate receivers have no energy storage capability. The legitimate receivers harvest the energy from the received signals by using a power splitting scheme to meet the circuit-power constraint for information decoding. The problem is formulated as a power allocation optimisation problem under the transmit power constraint and the energy harvesting constraint which is solvable. It is shown from simulations that the proposed optimal power allocation scheme outperforms the conventional uniform power allocation scheme.

In variable rate modulation, maximisation of data rate can be achieved either by fixing the symbol rate with an adaptive control of the constellation size, or by fixing the constellation size and changing the symbol duration. It is proved that for joint optimal control of both types, they should be used separately: the constellation size changes under a constant maximal symbol rate, and when these opportunities have been exhausted the symbol rate adaptation should be used.

A means of encoding and decoding data using wireless orbital angular momentum (OAM) modes is proposed and analysed. Source data symbols are used to select an OAM mode, which is generated using an 8-element circular array. A 2-element array is used to detect the mode by estimating the phase gradient of the received signal, and hence identifying the transmitted data symbol. The results are presented in terms of mode estimation error.