Department of Electrical Engineering

Highly c-axis-oriented zinc oxide (ZnO) thin films were deposited on Au electrodes by reactive radio frequency (RF) magnetron sputtering and their sputtering pressure on thin film bulk acoustic-wave resonator (FBAR) characteristics are presented. The evolution of the preferred orientation and the surface morphologies of the deposited ZnO films are investigated using X-ray diffraction, scanning electron microscopy, and atomic force microscopy measurement techniques. The result obtained in this study show that the ZnO films prepared using a lower sputtering pressure of 2 × 10−3 Torr have a strong c-axis orientation, promote smoother surface and higher resonance frequency. The experimental results demonstrate that the fabricated two-port FBAR using the optimum process parameters yields an effective electromechanical coupling constant ( $ k^{2}_{{{\text{eff}}}} $ ) of 2.8%, series quality factor (Q s) of 436, and a parallel quality factor (Q p) of 600.

The goal of hearing aids is to selectively attenuate acoustic signal based on the directional of sound arrival. Directional microphone or microphone arrays are used to achieve this goal. The longer distance of microphone pairs is a problem of the application on hearing aids. In this paper, a single directional microphone structure is proposed to solve this problem and can be used in other application. This directional microphone is constructed by the post CMOS MEMS process. The mechanism of this structure is originally inspired the by the ears of fly Ormia ochracea. In the result, the characteristic of this directional microphone is bi-direction pattern.

The method is described to integrate the 3x2 ladder type film bulk acoustic wave (FBAR) filter on CMOS chip. The modified Mason equivalent circuit model is used to simulate the characteristics of FBAR. The filter is designed by insertion loss method to meet the requirements. Low noise amplifier (LNA) has been designed and manufactured by the UMC 0.18um process. By the use of post CMOS process, the FBAR filter structure can be realized on CMOS chip. Finally, the mass loading frequency trimming method can adjust the center frequency of FBAR. The feasibility of integration can be proved by this method.

This paper proposes a method for using a pair of identical fiber Bragg gratings to measure the unbalanced strain factor in an isothermal-mechanical system. Cascading two identical fiber Bragg gratings, the unbalanced strain factor caused by the disturbance can be interrogated from the peak difference of the two reflecting Bragg wavelengths. If the mechanical system is in an isothermal environment, the thermal variations in the Bragg grating detection system can be automatically compensated. To verify the feasibility of the unbalanced strain detection using a pair of fiber Bragg gratings, the wall-thinning monitor in a pressure vessel system was investigated. The proposed technique is easy to install, immune to electro-magnetic and thermal interference, and highly sensitive in-time to remote monitoring for fiber optic smart structures. A finite element model (FEM) is used to simulate the wall thinning in a pressure vessel system. The minimum detectable unbalanced strain and cross-talk between two reflecting Bragg wavelengths were investigated.

In this paper, we used the theoretical calculation to simulate the response of SAW delay-line on quartz substrate and then compared to the experimental results. ST-cut quartz SAW delay-lines operating at 39.5MHz and 78.9MHz corresponding to 80 micron and 40 micron wavelength were developed. The simulation results agree with the experimental results shows that the simulation method can be used for quartz-based SAW delay-line as well as for SAW delayline applications; the sensors functioned correctly and can be used.

In this paper, the Film Bulk Acoustic wave Resonator (FBAR) and Band Pass Filter (BPF) have been designed, fabricated and characterized. And, a mask-less frequency trimming method is achieved through self assembling metal colloid electrode layers. The FBARs with AlN or ZnO piezoelectric layer and ladder type band pass filters have been realized by MEMS technique on wafer for verifying the frequency trimming mechanism. For bulk acoustic wave resonators, mass loading is an effective frequency trimming method, so mono-dispersed metal colloidal nanoparticles with size of 5±2 nm have been used as loading mass to trim resonant frequency. By the self-assembly of bi-functional group molecules, the mass of metal colloidal nanoparticles can be uniformly loaded layer by layer on the top electrode of tuned resonator. In the initial result, the 0.01% of resolution and 3% of tunablility can be achieved to meet the tight frequency specification.

Abstract A novel BInaural Spectro-Temporal (BIST) algorithm is proposed in this paper to increase the speech intelligibility in low or negative SNR noisy environments. The BIST algorithm consists of two modules. One is the spatial mask for receiving sound from the specific direction, and the other is the spectro-temporal modulation filter for noise reduction. Most speech enhancement algorithms are not applicable in harsh environments because the energy of speech is covered by the noise.

results show that the MGD covariance matrix coefficient of the auditory spectral centroid feature yields an accuracy of 83.87 % in detecting significant vascular access stenosis. Thus, the proposed method has significant potential for the applications of vascular access stenosis detection.

A novel BInaural Spectro-Temporal (BIST) algorithm is proposed in this paper to increase the speech intelligibility in low or negative SNR noisy environments. The BIST algorithm consists of two modules. One is the spatial mask for receiving sound from the specific direction, and the other is the spectro-temporal modulation filter for noise reduction. Most speech enhancement algorithms are not applicable in harsh environments because the energy of speech is covered by the noise. To increase the speech intelligibility in low or negative SNR noisy environments, a distinctive approach is proposed to solve this problem. First, the BIST algorithm takes binaural auditory processing as a spatial mask to separate the speech and noise according to their locations. Next, the modulation filter is applied to reduce the noise source in the scale-rate (spectro-temporal modulation) domain according to their different acoustic feature. It works like the spectro-temporal receptive field (STRF) which is the perception response of human auditory cortex. The experimental results demonstrate that the proposed BIST speech enhancement algorithm can improve 20% from the noisy speech at SNR-10dB.

This study presents a computer-assisted auscultation approach for patent ductus arteriosus (PDA) detection. PDA is a frequent congenital heart disease in neonates. The early detection of the PDA murmur from a neonate's heart sound is important to avoid the development of severe respiratory diseases in newborns. Computer-assisted auscultation has the potential to become a heart disease screening tool because it can improve the sensitivity and specificity of heart murmur detection. This study proposes an auditory time-frequency representation method (auditory spectrogram) to demonstrate the dynamic and non-stationary characteristics of heart murmurs. The auditory spectrogram is obtained using the proposed cochlear wavelet transform, which mimics the auditory response of the cochlea in the human auditory system. The auditory spectrogram can represent the heart sounds in the time-frequency domain to assist cardiac auscultation. Furthermore, the spectrum flux and spectrum centroid are used to characterize the configuration of the auditory spectrogram. The phonocardiogram data of 18 patients, including normal heart sounds, innocent murmurs, and PDA murmurs, are used in the blind test for algorithm effectiveness assessment. The results demonstrate that the proposed computer-assisted auscultation method can achieve a high sensitivity of 100 % and a specificity of 91.67 % for PDA detection. The developed method is robust, cost-effective, and convenient, making it effective for the enhancement of the accuracy of PDA detection.

An auditory-based feature extraction algorithm naming the Basilar-membrane Frequency-band Cepstral Coefficient (BFCC) is proposed to increase the robustness for automatic speech recognition. Compared to Fourier spectrogram based of the Mel-Frequency Cepstral Coefficient (MFCC) method, the proposed BFCC method engages an auditory spectrogram based on a gammachirp wavelet transform to simulate the auditory response of human inner ear to improve the noise immunity. In addition, the Hidden Markov Model (HMM) is used for evaluating the proposed BFCC in phases of training and testing purposes conducted by AURORA-2 corpus with different Signal-to-Noise Ratios (SNRs) degrees of datasets. The experimental results indicate the proposed BFCC, compared with MFCC, Gammatone Wavelet Cepstral Coefficient (GWCC), and Gammatone Frequency Cepstral Coefficient (GFCC), improves the speech recognition rate by 13%, 17%, and 0.5% respectively, on average given speech samples with SNRs ranging from-5 to 20 dB.

We show the multiple description scalar quantizer (MDSQ) [l], when used over a slow fading Rayleigh channel, results in a smaller interleaving delay than traditional channel code-based approaches and substantially outperforms a system using a maximum ratio combiner. We consider a continuous alphabet source without memory over a slow Rayleigh fading channel. Specifically, it is initially assumed the Rayleigh parameter remains constant over a specific time interval and the demodulator has available perfect estimates of the Rayleigh parameter. An analysis of the problem is presented and theoretical justification is provided for using the MDSQ. For a Gaussian source without memory, performance comparisons are presented against a traditional maximum ratio combiner (MRC)-based system as well as against channel code-based systems. It is shown that the performance of the MDSQ-based system dominates the MRC-based system at all channel signal-to-noise ratios and is superior to channel code-based systems when the interleaving delay is constrained. Simulation results for a mobile radio channel model at 15 mph indicate a reduction in the interleaving delay by at least a factor of three.

We present a simple and fast algorithm for image noise estimation. The input image is assumed to be corrupted by additive zero mean Gaussian noise. To exclude structures or details from contributing to the noise variance estimation, a simple edge detection algorithm using first-order gradients is applied first. Then a Laplacian operator followed by an averaging over the whole image will provide very accurate noise variance estimation. There is only one parameter which is self-determined and adaptive to the image contents. Simulation results show that the proposed algorithm performs well for different types of images over a large range of noise variances. Performance comparisons against other approaches are also provided.

Machine-to-machine (M2M) communication is an emerging technology to provide ubiquitous connectivity among devices without human intervention. The cellular networks are considered a ready-to-use infrastructure to implement M2M communications. However, M2M communications over cellular pose significant challenges to cellular networks due to different data transactions, diverse applications, and a large number of connections. To support such a large number of devices, M2M system architecture should be extremely power and spectrum efficient. In this paper, we provide a comprehensive survey on M2M communications in the context of the Third-Generation Partnership Project (3GPP) Long-Term Evolution (LTE) and Long-Term Evolution-Advanced (LTE-A). More specifically, this paper presents architectural enhancements for providing M2M services in 3GPP LTE/LTE-A networks and reviews the features and requirements of M2M applications. In addition, the signal overheads and various quality-of-service (QoS) requirements in M2M communications also deserve our attention. We address M2M challenges over 3GPP LTE/LTE-A and also identify the issues on diverse random access overload control to avoid congestion caused by random channel access of M2M devices. Different application scenarios are considered to illustrate futuristic M2M applications. Finally, we present possible enabling technologies and point out the directions for M2M communications research.

—Providing diverse and strict quality of service (QoS) guarantees is one of the most important requirements in M2M communications, which particularly need for appropriate resource allocation for a large number of M2M devices. To efficiently allocate resource blocks (RBs) for M2M devices while satisfying QoS requirements, we propose group based M2M communications , in which M2M devices are clustered based on their wireless transmission protocols, their QoS characteristics and requirements. To perform joint RB and power allocation in SC-FDMA based LTE-A networks, we formulate a sum-throughput maximization problem, while respecting all the constraints associated with SC-FDMA scheme as well as QoS requirements in M2M devices. The constraints in uplink SC-FDMA air interface in LTE-A networks complicate the resource allocation problem. We solve the resource allocation problem by first transforming it into a binary integer programming (BIP) problem, and then formulate a dual problem using the Lagrange duality theory. Numerical results show that the proposed algorithm outperforms traditional Greedy algorithm in terms of throughput maximization while satisfying QoS requirements, and its performance is close to the optimal design.

Global connectivity, low latency, and ready‐to‐use infrastructure of next generation wireless (NGW) networks provide a platform for machine‐to‐machine (M2M) communications on a large scale. However, M2M communications over NGW networks pose significant challenges because of different data rates, diverse applications , and a large number of connections. In this paper, we address M2M challenges over NGW networks, and in particular, we focus on random access overload issue and diverse quality‐of‐service (QoS) requirements to enable M2M communications in the context of NGW networks. To enable massive M2M access while QoS guarantees, we propose group‐based M2M communications on the basis of identical transmission protocols and QoS requirements. Furthermore, to guarantee low energy consumption for M2M devices in the same group, we propose a decentralized group‐head selection scheme. In addition, a solution is proposed by using an effective capacity concept to provide QoS guarantees for M2M devices with a strict time constraint. A new random access approach based on different lengths super orthogonal codes is proposed to ease massive random access challenges with provisioning diverse QoS requirements of M2M communications in heterogeneous NGW networks.

Enabling cellular connectivity for drones introduces a wide set of challenges and opportunities. Communication of cellular-connected drones is influenced by 3-dimensional mobility and line-of-sight channel characteristics which results in higher number of handovers with increasing altitude. Our cell planning simulations in coexistence of aerial and terrestrial users indicate that the severe interference from drones to base stations is a major challenge for uplink communications of terrestrial users. Here, we first present the major challenges in co-existence of terrestrial and drone communications by considering real geographical network data for Stockholm. Then, we derive analytical models for the key performance indicators (KPIs), including communications delay and interference over cellular networks, and formulate the handover and radio resource management (H-RRM) optimization problem. Afterwards, we transform this problem into a machine learning problem, and propose a deep reinf...

Cellular connectivity for drones comes with a wide set of challenges as well as opportunities. Communication of cellular-connected drones is influenced by 3-dimensional mobility and line-of-sight channel characteristics which results in higher number of handovers with increasing altitude. Our cell planning simulations in coexistence of aerial and terrestrial users indicate that the severe interference from drones to base stations is a major challenge for uplink communications of terrestrial users. Here, we first present the major challenges in co-existence of terrestrial and drone communications by considering real geographical network data for Stockholm. Then, we derive analytical models for the key performance indicators (KPIs), including communications delay and interference over cellular networks, and formulate the handover and radio resource management (H-RRM) optimization problem. Afterwards, we transform this problem into a machine learning problem, and propose a deep reinfor...

The main target of this paper is to propose a preferred set of features from a cellular network for using as predictors to do the classification between the flying drone User Equipments (UEs) and regular UEs for different Machine Learning (ML) models. Furthermore, the target is to study four different machine learning models i.e. Decision Tree (DT), Logistic Regression (LR). Discriminant Analysis (DA) and K-Nearest Neighbour (KNN) in this paper, and evaluate/compare their performance in terms of identifying the flying drone UE using three performance metrics i.e. True Positive Rate (TPR), False Positive Rate (FPR) and area under Receiver Operating Characteristic (ROC) curve. The simulations are performed using an agreed 3GPP scenario, and a MATLAB machine learning tool box. All considered ML models provide high drone detection probability for drones flying at 60 m and above height. However, the true drone detection probability degrades for drones at lower altitude. Whereas, the fine DT method and the coarse KNN model performs relatively better compared with LR and DA at low altitude, and therefore can be considered as a preferable choice for a drone classification problem.