Recently, a next-generation heterodyne mixer detector---a hot electron bolometer (HEB) mixer employing a superconducting microbridge---has gradually opened up terahertz-band astronomy. The surrounding state-of-the-art technologies including fabrication processes, 4 K cryostats, cryogenic low-noise amplifiers, local oscillator sources, micromachining techniques, and spectrometers, as well as the HEB mixers, have played a valuable role in the development of super-low-noise heterodyne spectroscopy systems for the terahertz band. The current developmental status of terahertz-band HEB mixer receivers and their applications for spectroscopy and astronomy with ground-based, airborne, and satellite telescopes are presented.

To immobilize cytochrome $c$ (cyt.,$c$) on ITO electrode with keeping its direct electron transfer (DET) activity, 10-carboxydecylphosphonic acid (10-CDPA) self-assembled monolayer (SAM) film was formed on ITO electrode. After 100 times washing process with exchanging phosphate buffer saline solution in the cell to fresh one, extit{in situ} slab optical waveguide (SOWG) absorption spectral measurement proved that about 80% of cyt.,$c$ immobilized on 10-CDPA modified ITO electrode was adsorbed on ITO electrode. Additionally SOWG spectral change of cyt.,$c$ between oxidized and reduced forms was observed with setting the ITO electrode potential at 0.3 and $-$0.3,V vs. Ag/AgCl, respectively showing DET reaction between cyt.,$c$ and ITO electrode occurred. About 30% of a monolayer coverage was estimated from the coulomb amount in the surface area of oxidation and reduction peaks on cyclic voltammetry (CV) data. CV peak current maintained 84% for ITO electrode modified with 10-CDPA SAM film after 60,min continuous scan with 0.1,V/sec from 0.3 and $-$0.3,V vs. Ag/AgCl.

A portable biofuel cell with an anode modified with glucose oxidase/ferrocene and a cathode modified with bilirubin oxidase was fabricated on a flexible polyimide substrate. The anode and the cathode with an area of 3$ imes$10,mm$^2$ were separated with a gap of 1,mm. Solidified glucose biofuel units were prepared by solidifying 50--200,mM glucose aqueous solution containing 0.5--2.5% agarose. The influences of the biofuel volume and glucose concentrations on power generation were investigated. The maximum power density was almost independent on the agarose concentration, and it continuously decreased as time clasped possibly due to the consumption of glucose and/or release of the enzymes from the electrodes. The maximum power and power density were 0.32 $mu $W and 1.08,$mu $W/cm$^2$ at 0.17,V, respectively when the solidified hydrogel biofuel unit with a dimension of 13$ imes$24$ imes$4 mm$^3$ containing the 100,mM glucose aqueous solution and 2.5, wt% agarose was used. The biofuel cell continued to deliver the power density over 0.5,$mu $W/cm$^2$ for more than 1,h.

We have investigated both the film thickness and surface potential of organic semiconductors deposited on two kinds of electrodes by the simultaneous observation with the dynamic force microscopy (DFM)/Kelvin-probe force microscope (KFM). To clarify the interfacial properties of organic semiconductor, we fabricated samples that imitated the organic light emitting diode (OLED) structure by depositing bis [$N,N '$-(1-naphthyl)-$N,N '$-phenyl] benzidine ($alpha$-NPD) and tris (8-hydroxyquinolinato) aluminum (Alq$_{3}$), respectively, on indium-tin-oxide (ITO) as anode and aluminum (Al) as cathode by the vacuum evaporation deposition using intersecting metal shadow masks. This deposition technique enables us to fabricate four different areas in the same substrate. The crossover area of the deposited thin films were measured by the DFM/KFM, the energy band diagrams were depicted and we considered that the charge behavior of the organic semiconductor depended on the material and the structure.

We report a real time method to monitor the chemical reaction in microdroplets, which contain an organic dye, 5(6)-carboxynaphthofluorescein and a CdSe/ZnS quantum dot using fluorescence spectra. Especially, the relationship between the droplet size and the reaction rate of the two reagents was investigated by changing an injection speed.

Step-edge vertical channel organic field-effect transistors (SVC-OFETs) with a very short channel have been fabricated by a novel selective electrospray deposition (SESD) method. We propose the SESD method for the fabrication of SVC-OFETs based on a 6,13-bis(triisopropyl-silylethynyl) pentacene (TIPS-pentacene) semiconductor layer formed by SESD. In the SESD method, an electric field is applied between the nozzle and selective patterned electrodes on a substrate. We demonstrated that the solution accumulates on the selected electrode pattern by controlling the voltage applied to the electrode.

A multi-port device is characterized using measurement results of a two-port Vector Network Analyzer (VNA) with four different structures. The loads used as terminations are open-, or short-circuited transmission lines (TLs), which are characterized along with Ground-Signal-Ground pads based on L-2L de-embedding method. A new characterization method for a four-port device is introduced along with its theory. The method is validated using simulation and measurement results. The characterized four-port device is a Crossing Transmission Line (CTL), mainly used for over-pass or under-pass of RF signals. Four measurement results are used to characterize the CTL. The S-parameter response of the CTL is found. To compare the results, reconstructed responses compared with the measurements. Results show good agreement between the measured and modeled results from 1 GHz to 110 GHz.

In this paper, we propose a new cross-layer scheme Cooperation between channel Access control and TCP Rate Adaptation (CATRA) aiming to manage TCP flow contention in multi-hop ad hoc networks. CATRA scheme collects useful information from MAC and physical layers to estimate channel utilization of the station. Based on this information, we adjust Contention Window (CW) size to control the contention between stations. It can also achieve fair channel access for fair channel access of each station and the efficient spatial channel usage. Moreover, the fair value of bandwidth allocation for each flow is calculated and sent to the Transport layer. Then, we adjust the sending rate of TCP flow to solve the contention between flows and the throughput of each flow becomes fairer. The performance of CATRA is examined on various multi-hop network topologies by using Network Simulator (NS-2).

Support of incoming traffic differentiation and Quality of Service (QoS) assurance is very important for the development of high performance packet switches, capable of separating traffic flows. In our previous paper, we proposed the implementation of two buffers at each crosspoint of a crossbar fabric that leads to the Dual Crosspoint Queued (DCQ) switch. Inside DCQ switch, one buffer is used to store the real-time traffic and the other for the non-real-time traffic. We also showed that the static priority algorithms can provide the QoS only for the real-time traffic due to their greedy nature that gives the absolute priority to that type of traffic. In order to overcome this problem, in our paper we propose the DCQ switch with the Largest Weighted Occupancy First scheduling algorithm that provides the desired QoS support for both traffic flows. Detailed analysis of the simulation results confirms the validity of proposed solution.

This paper proposes to obtain high-level, domain-robust representations for cross-view face recognition. Specially, we introduce Convolutional Deep Belief Networks (CDBN) as the feature learning model, and an CDBN based interpolating path between the source and target views is built to model the correlation of cross-view data. The promising results outperform other state-of-the-art methods.

As VLSI process node continue to shrink, chemical mechanical planarization (CMP) process for copper interconnect has become an essential technique for enabling many-layer interconnection. Recently, Edge-over-Erosion error (EoE-error), which originates from overpolishing and could cause yield loss, is observed in various CMP processes, while its mechanism is still unclear. To predict these errors, we propose an EoE-error prediction method that exploits machine learning algorithms. The proposed method consists of (1) error analysis stage, (2) layout parameter extraction stage, (3) model construction stage and (4) prediction stage. In the error analysis and parameter extraction stages, we analyze test chips and identify layout parameters which have an impact on EoE phenomenon. In the model construction stage, we construct a prediction model using the proposed multi-level machine learning method, and do predictions for designed layouts in the prediction stage. Experimental results show that the proposed method attained 2.7∼19.2% accuracy improvement of EoE-error prediction and 0.8∼10.1% improvement of non-EoE-error prediction compared with general machine learning methods. The proposed method makes it possible to prevent unexpected yield loss by recognizing EoE-errors before manufacturing.

Using “clean-slate approach” to redesign the Internet has attracted considerable attention. ZNA (Z Network Architecture) is one of clean-slate network architectures based on the layered model. The major features of ZNA are as follows: (1) introducing the session layer to provide the applications with sophisticated communication services, (2) employing inter-node cross-layer cooperation to adapt to the dynamically changing network conditions, (3) splitting the node identifier and the node locator for mobility, multi-homing, and heterogeneity of network layer protocols, (4) splitting the data plane and the control plane for high manageability, and (5) introducing a recursive layered model to support network virtualization. This paper focuses on the first three topics as well as the basic design of ZNA.

In this paper, we use neural networks (NNs) for cross-dialectal (Mandarin-Shanghainese) voice conversion using a bi-dialectal speakers' recordings. This system employs a nonlinear mapping function, which is trained by parallel mandarin features of source and target speakers, to convert source speaker's Shanghainese features to those of target speaker. This study investigates three training aspects: a) Frequency warping, which is supposed to be language independent; b) Pre-training, which drives weights to a better starting point than random initialization or be regarded as unsupervised feature learning; and c) Sequence training, which minimizes sequence-level errors and matches objectives used in training and converting. Experimental results show that the performance of cross-dialectal voice conversion is close to that of intra-dialectal. This benefit is likely from the strong learning capabilities of NNs, e.g., exploiting feature correlations between fundamental frequency (F0) and spectrum. The objective measures: log spectral distortion (LSD) and root mean squared error (RMSE) of F0, both show that pre-training and sequence training outperform the frame-level mean square error (MSE) training. The naturalness of the converted Shanghainese speech and the similarity between converted Shanghainese speech and target Mandarin speech are significantly improved.

In this paper, the H2/H∞ control problem for a class of stochastic discrete-time linear systems with state-, control-, and external-disturbance-dependent noise or (x, u, v)-dependent noise involving multiple decision makers is investigated. It is shown that the conditions for the existence of a strategy are given by the solvability of cross-coupled stochastic algebraic Riccati equations (CSAREs). Some algorithms for solving these equations are discussed. Moreover, weakly-coupled large-scale stochastic systems are considered as an important application, and some illustrative examples are provided to demonstrate the effectiveness of the proposed decision strategies.

This paper provides a prototype neural prosthesis system dedicated to restoring continence and micturition function for patients with lower urinary tract diseases, such as detrusor hyperreflexia and detrusor-sphincter dyssynergia. This system consists of an ultra low-noise electroneurogram (ENG) signal recording module, a bi-phasic electrical stimulator module and a control unit for closed-loop bladder monitoring and controlling. In order to record extremely weak ENG signal from extradural sacral nerve roots, the system provides a programmable gain from 80 dB to 117 dB. By combining of advantages of commercial-off-the-shelf (COTS) electronics and custom designed IC, the recording front-end acquires a fairly low input-referred noise (IRN) of 0.69 μVrms under 300 Hz to 3 kHz and high area-efficiency. An on-chip multi-steps single slope analog-to-digital converter (ADC) is used to digitize the ENG signals at sampling rate of 10 kSPS and achieves an effective number of bits (ENOB) of 12.5. A bi-phasic current stimulus generator with wide voltage supply range (±0.9 V to ±12.5 V) and variable output current amplitude (0-500 μA) is introduced to overcome patient-depended impedance between electrode and tissue electrolyte. The total power consumption of the entire system is 5.61 mW. Recording and stimulation function of this system is switched by control unit with time division multiplexing strategy. The functionality of this proposed prototype system has been successfully verified through in-vivo experiments from dogs extradural sacral nerve roots.

This paper proposes a compact three-mode H-shaped resonator bandpass filter fed by antiparallel coupled input/output lines. To investigate the resonant behavior of the H-shaped resonator, even/odd-mode resonance conditions of the resonator are first derived analytically. The multimode resonances of the H-shaped resonator filter are modeled by a multipath circuit formed with resonance paths. Moreover, a direct source/load coupling path is connected in parallel, of which the value shows a frequency dependency because of the antiparallel coupled feeding lines, thereby generating four transmission zeros (TZs) greater than the number of a theoretical limitation. The H-shaped resonator bandpass filter is synthesized, developed, and tested, showing a third-order passband response with four TZs located near the passband, and a wide stopband property.

This paper presents the design and radiation properties of a linearly polarized radial line microstrip antenna array (RL-MSAA) with U-slot circular microstrip antennas. A circular microstrip antenna (C-MSA) with U-shaped slot is used as a radiation element of the RL-MSAA. Radiation phase of the U-slot C-MSA is controlled by tuning the radius of the C-MSA and dimensions of the U-slot on the C-MSA; therefore, the desired phase distribution of the RL-MSAA can be realized. In this paper, a linearly polarized RL-MSAA with three concentric rows of C-MSAs at a spacing of 0.65 wavelengths is designed for 12GHz operation. In order to realize uniform phase distribution, the U-slot C-MSAs are arranged for inner two rows and normal C-MSAs are arranged for the termination row. Validity of the linearly polarized RL-MSAA with the U-slot C-MSAs for radiation phase control is demonstrated by simulation and measurement.

Wireless body area networks (WBANs) have to work with low power and long lifetime to satisfy human biological safety requirements in e-health; therefore extremely low energy consumption is significant for WBANs. IEEE 802.15.6 standard has been published for wearable and implanted applications which provide communication technology requirements in WBANs. In this paper, the cross-layering optimization methodology is used to minimize the network energy consumption. Both the priority strategy and sleep mechanism in IEEE802.15.6 are considered. Macroscopic sleep model based on WBAN traffic priority and microscopic sleep model based on MAC structure are proposed. Then the network energy consumption optimization problem is solved by Lagrange dual method, the master problem are vertically decomposed into two sub problems in MAC and transport layers which are dealt with gradient method. Finally, a solution including self-adaption sleep mechanism and node rate controlling is proposed. The results of this paper indicate that the algorithm converges quickly and reduces the network energy consumption remarkably.

Binaural reproduction is one of the promising approaches to present a highly realistic virtual auditory space to a listener. Generally, binaural signals are reproduced using a set of headphones that leads to a simple implementation of such a system. In contrast, binaural signals can be presented to a listener using a technique called “transaural reproduction” which employs a few loudspeakers with crosstalk cancellation for compensating acoustic transmissions from the loudspeakers to both ears of the listener. The major advantage of transaural reproduction is that a listener is able to experience binaural reproduction without wearing any device. This leads to a more natural listening environment. However, in transaural reproduction, the listener is required to be still within a very narrow sweet spot because the crosstalk canceller is very sensitive to the listener's head position and orientation. To solve this problem, dynamic transaural systems have been developed by utilizing contact type head tracking. This paper introduces the development of a dynamic transaural system with non-contact head tracking which releases the listener from any attachment, thereby preserving the advantage of transaural reproduction. Experimental results revealed that sound images presented in the horizontal and median planes were localized more accurately when the system tracked the listener's head rotation than when the listeners did not rotate their heads or when the system did not track the listener's head rotation. These results demonstrate that the system works effectively and correctly with the listener's head rotation.

A major challenge for speech emotion recognition is that when the training and deployment conditions do not use the same speech corpus, the recognition rates will obviously drop. Transfer learning, which has successfully addressed the cross-domain classification or recognition problem, is presented for cross-corpus speech emotion recognition. First, by using the maximum mean discrepancy embedding (MMDE) optimization and dimension reduction algorithms, two close low-dimensional feature spaces are obtained for source and target speech corpora, respectively. Then, a classifier function is trained using the learned low-dimensional features in the labeled source corpus, and directly applied to the unlabeled target corpus for emotion label recognition. Experimental results demonstrate that the transfer learning method can significantly outperform the traditional automatic recognition technique for cross-corpus speech emotion recognition.