Resonant tunneling diodes (RTDs) have the potential for compact and coherent terahertz (THz) sources operating at room temperature. In this paper, recent results of THz oscillators with RTDs are described. A fundamental oscillation frequency up to 831GHz was achieved with RTD having high available current density and low capacitance. By the structure reducing the transit time, the frequency further increased to 1.04THz. This is the first achievement of a fundamental oscillation above 1THz in room-temperature electronic single oscillators. The output power of 400µW at 550GHz was obtained in a single oscillator by the offset-fed slot antenna. Coherent power combining with multi-element array was observed. The spectral linewidth, frequency change with bias voltage, and direct modulation were also described.
A number of technical breakthroughs in electronics and photonics made since the early 1990s have started to bring terahertz (THz)-wave technologies from laboratory demonstrators to industrial applications such as non-destructive testing, security, medicine, communications, etc. This paper overviews the latest progress in THz-wave technologies in terms of components such as sources and detectors, and system applications, and discusses future challenges towards market developments.
This paper presents a new method to extract linear-scale perceptual feature as a subsitute of MFCCs for highband (3.4kHz∼) in Speech Bandwidth Extensions(BWE). The feature extraction method is based on the mel-scale constrained Nonnegative Matrix Factorization(NMF), which decompose linear-scale log spectrum into a linear combination of mel-scale latent variables. While MFCCs parametrization contains non-invertible procedures, suggested feature is represented in linear-scale and proper to recover the highband time-domain speech. Experiment results report that suggested feature shows better instrumental performance with narrowband MFCCs than real cepstrum without additional computation.
Focused word spotting makes a contribution to keyword extraction and speech understanding in spoken Korean. The average and variance of fundamental frequency were expected to be important factors to detect verbal focus of a speaker. From the experiment of focused word spotting in each sentence by statistically modeling those factors, I achieved a spotting accuracy of 84%.
Emotion of speech degrades the performance of Automatic Speech Recognition (ASR) systems. With the aim of enhancing the emotional speech recognition accuracy, the effects of formant frequencies and their slopes on improving the performance are investigated in this paper. For this purpose, the formant frequencies are neutralized using hybrid of Dynamic Time Warping (DTW) and Multi-Layer Perceptron (MLP) neural networks. Each one of the neutralized formant frequencies is considered as a supplementary feature and used in Hidden Markov Model (HMM)-based ASR. Experimental results show that by using the slope of neutralized formant frequency features, the recognition rate in happiness and anger states is improved by at most 2.1% and 3.6%, respectively.
The number of sets, block size, and associativity determine processor's cache configurations. Particularly in embedded systems, their cache configuration can be optimized since their target applications are much limited. Recently, the CRCB method has been proposed for LRU-based (Least Recently Used-based) cache configuration simulation, which can calculate cache hit/miss counts accurately and very fast changing the three parameters. However many recent processors use FIFO-based (First-In-First-Out-based) caches instead of LRU-based caches due to the viewpoints of their hardware costs. In this paper, we propose a speeding-up cache configuration simulation method for embedded applications that uses FIFO as a cache replacement policy. We first prove several properties for FIFO-based caches and then propose a simulation method that can process two or more FIFO-based cache configurations with different cache associativities simultaneously. Experimental results show that our proposed method can obtain accurate cache hits/misses and runs up to 32% faster than the conventional simulators.
We present the design of a compact dual-band bandpass filter using open-loop resonator, which connected together with mixed coupling, loaded by T- shape stub and open stubs. Open stubs are loaded to achieve high performance and the tapped line feed structure is used in input-output port. By adjusting the tapping position, stubs length and the impedance ratio of T- shape stub, the filter performance can be controlled. The T shape stubs are used to suppress the harmonics above the second band. The insertion losses are 0.2dB and 0.9dB at 2.7 and 5GHz respectively. Above the second pass band minimum attenuation is -20dB from 5.3 to 20GHz, so it can be used for ultra wide stopband applications. The measurement results have shown a good agreement with the simulation.
A low power receiver RF front-end operating in the 5GHz band according to the IEEE 802.11a WLAN standard is reported. This is based on the dual conversion architecture for the fully integrated receiver, in order to relax the problem of the direct conversion receiver. In the RF front-end, the power consumption is greatly reduced by using the dual conversion current-driven passive mixer with linear transconductor stage. The receiver RF front-end is fabricated with 0.13µm CMOS technology. An IIP3 of -15.8dBm with a gain of 37.2dB and a noise figure of 7.1dB are obtained at 16.8mW power consumption.
Currently, Intelligent Transportation System (ITS) is a promising technique that can efficiently manage the serious traffic problems in metropolises. It adopts the mobile communication techniques to transmit traffic information in many applications such as the inter-vehicle communications (IVC) and the vehicle-road communications (VRC). The longer distance of the information is transmitted, the more safety can be achieved for the drivers because they will have enough time to avoid or reduce the damage. In this paper, combined with a multi-ray shadowing model, the range extension capability by using the two receiving antennas is evaluated.
The paper aims the electromagnetic simulation of a synchronous impedance transformer using a combination of the finite difference time domain FDTD and the method of lines MoL. The study demonstrates that the hybrid technique called FDTD-MoL can be used for the characterization of passive microstrip circuits. The capability to combine different methods of electromagnetic simulation provides a powerful tool for interpreting the behavior of microwave circuits. The hybrid FDTD-MoL method can characterize microstrip planar structures and is comparable with previous case studies of electromagnetic simulation.
A 140GHz CMOS wideband amplifier is proposed with the low group delay variation required to achieve a high-speed D-band wireless receiver. The amplifier is fabricated by the standard 1P12M 65nm CMOS process. From measurement, the gain is 10dB with a group delay variation of 10.2ps. The 0.1dB bandwidth, used as the figure of merit of the gain flatness, is 12GHz, whereas the generally used 3dB bandwidth is 27.6GHz. The power consumption is 57.1mW with a supply voltage of 1.2V.
A new method for the dynamic range enhancement of the PIM-producing patch antenna with diode is proposed, and its effectiveness is verified by experiment. In 2GHz-band experiment, we shows that the size reduction of the diode-mounted hole has an equivalent effect of the diode-rotation in the previous work, and contributes to the linearity improvement in the high power excitation. Through the testing result of 800MHz-band antenna, it is demonstrated that good linearity is obtained up to 43dBm excitation in swept-power test when Dh=4.6mm. The frequency characteristics of the employed diode are also verified through a frequency-swept test. The resultant PIM-level variation is less than 2dB, which is negligibly small in comparison with the difference between 800MHz- and 2GHz-antennas.