Search

A $X-/K_{\mathrm{u}}$ -band low-noise amplifier (LNA) with quad flat no-leads (QFN) packaging for the satellite communication application is proposed in this work. The LNA is able to support the reception of dual horizontally (H) and vertically (V) polarized signals, increasing the channel capacity. The parasitic effect of the QFN package is captured by a 3D electromagnetic simulator and is then co-designed with the LNA to achieve simultaneous noise and impedance matching at the band of interest. Realized in a $0.15-\mu \mathrm{m}$ GaAs pHEMT technology, the proposed QFN-packaged LNA can provide power gain of 20.6 and 21.6 dB while having 3-dB bandwidth from 10.7 to 13.1 GHz and minimum noise figure (NF) of 1.3 dB for the H- and V-polarization channels, respectively. The power consumption is only 43.2 mW from a 1-V supply.

Ultrafast optical pulse generation plays important role in pump/probe measurement for the fastest dynamics study in atoms, molecules and materials. Among all of nonlinear effects to generate short pulses, high-order harmonic generation (HHG) opens a path towards the production of the shortest light pulses, with durations in the order of the attosecond time scale. The generation mechanism of HHG is easily understood by the semi-classical three-step model: an electron is ionized and accelerated by an intense driving electrical field; after half-cycle of the driving pulse, the electron is pulled back, re-colliding with the parent ion [1,2]. The kinetic energy accumulated during propagation in the continuum is released as high-energy -or high-frequency-radiation, ranging from the extreme ultraviolet (EUV) to the soft X-rays. Due to the nature of the HHG process, attosecond bursts are emitted every half cycle of the infrared (IR) laser field, resulting in the generation of an attosecond pulse train when multi­cycle driving pulses are used. This fact deprives the attosecond time scale of the HHG radiation, since the envelope width of the APT is usually in the range of several femtoseconds. Many approaches have demonstrated the production of isolated attosecond pulses, giving access to study attosecond dynamics [3,4]. However, those methods are limited to linearly polarized fields. When an elliptically polarized driving laser field is applied, the electronic wavepacket is driven away from the parent ion, decreasing the probability of recombination and thus significantly dropping the harmonic yield. The availability of circularly polarized or elliptically polarized attosecond pulses is particularly interesting for the study of dynamics in circular dichromic metrology of magnetic materials, chiral molecules and electronic spin of matter waves [5,6]. The great challenge of HHG still remains in this community: how to control the polarization state of attosecond pulses. Though large facilities, such as synchrotrons or x-ray free electron lasers (XFEL), provide alternative access to circularly polarized EUV or X-ray radiation with high brightness (by means of specially designed undulator systems [7]), their access is limited. In addition, the shortest pulse duration reported to date in XFELs is in the order of tens of femtoseconds, which imposes a limitation for investigating fast dynamics of electron spin motion. Therefore, polarization control of attosecond pulse on tabletop EUV light source becomes essential. In this talk, I will review the recent progress to produce elliptically or circularly polarized attosecond pulses. In particular, we show that today complete polarization control of HHG is available, through adjusting the ellipticity of two input driving pulses with counter-rotating helicity in a non-collinear geometry [8]. This solves the great dilemma of polarization control in HHG, paving the way towards attosecond metrology in chiral systems and magnetic materials.

The main purpose of this research is to detect and classify the solar power system which were displayed, abandoned or destroyed in MingDao university and along with the vertical axis wind turbine design as an electrical power generation system. The whole system is built at the lakeside which can supply power for night light and illumination lasts 36 hours. The system is an independent power producing system. Adopt the Arduino and related electronic sensor module to collect data. It will store massive power production data of the system to the SD card and build a database. The data can be used as power dispatching, monitoring and analysis. This research is a pioneer project, in the future, Arduino module data will be an important reference for the 12 hectares of green energy power to build in school.

Bi-directional free-space optical communication with bidirectional 2x4x10 Gb/s wavelength division multiplexing channel signals is demonstrated. The environmental factor effects including thermally induced non-uniform air index and rainfall on the FSO performance are investigated.

Galled trees are a class of tree-like phylogenetic networks in which the loops are not overlapping with each other. They are popularly used among biologists to represent the evolutionary history between a set of species. In this paper, we propose a compact encoding for the structure of a galled tree, and show that with our encoding, the tree containment problem on a galled tree can be solved in optimal time.

A cluster validity index (CVI) classification method is applied to enhance the performance of existing Multiple-Attribute Decision Making (MADM) method. This paper proposed index-based method is called the FRM-index method which combined Fuzzy Set (FS), Rough Set (RS), and a cluster validity index function. The effectiveness of the proposed FRM-index method is evaluated by comparing the classification results obtained for the relating UCI datasets using a statistical approach. Overall, the results show that the proposed method not only provides a more reliable basis for the extraction of decisionmaking rules for multi-output datasets, but also fills out the uncertainty and facilitates an effective MADM built.

It is said that a picture is worth a thousand words. Thus, there are various ways to describe an image, especially in aesthetic quality analysis. Although aesthetic quality assessment has generated a great deal of interest in the last decade, most studies focus on providing a quality rating of good or bad for an image. In this work, we extend the task to produce captions related to photo aesthetics and/or photography skills. To the best of our knowledge, this is the first study that deals with aesthetics captioning instead of AQ scoring. In contrast to common image captioning tasks that depict the objects or their relations in a picture, our approach can select a particular aesthetics aspect and generate captions with respect to the aspect chosen. Meanwhile, the proposed aspect-fusion method further uses an attention mechanism to generate more abundant aesthetics captions. We also introduce a new dataset for aesthetics captioning called the Photo Critique Captioning Dataset (PCCD), which contains pair-wise image-comment data from professional photographers. The results of experiments on PCCD demonstrate that our approaches outperform existing methods for generating aesthetic-oriented captions for images.

A multi-mode WPAN transceiver implemented in an SoC is presented. Fabricated in 40-nm CMOS, the chip supports IEEE 802.15.4 and all modes of Bluetooth. Consuming 7.8 mW from a 1.2-V supply, the receiver has a sensitivity of −104, −98, −95/−94/−88 dBm in 802.15.4, BLE and Bluetooth BDR/EDR2/EDR3, respectively. The transmitter has a power consumption of 10 mW to deliver 0 dBm output in the constant envelope mode and 14.5 mW in BT EDR.

As high-resolution fingerprint images are becoming more common, the pores have been found to be one of the promising candidates in improving the performance of automated fingerprint identification systems (AFIS). This paper proposes a deep learning approach towards pore extraction. It exploits the feature learning and classification capability of convolutional neural networks (CNNs) to detect pores on fingerprints. Besides, this paper also presents a unique affine Fourier moment-matching (AFMM) method of matching and fusing the scores obtained for three different fingerprint features to deal with both local and global linear distortions. Combining the two aforementioned contributions, an EER of 3.66% can be observed from the experimental results.

Aesthetic quality estimation of an image is a challenging task. In this paper, we introduce a deep CNN approach to tackle this problem. We adopt the sate-of-the-art object-recognition CNN as our baseline model, and adapt it for handling several high-level attributes. The networks capable of dealing with these high-level concepts are then fused by a learned logical connector for predicting the aesthetic rating. Results on the standard benchmark shows the effectiveness of our approach.

With the flourishing development of computer technology, image duplication can be carried out in a convenient way. In order to protect intellectual property rights, experts are devoted to many anti-counterfeiting technologies. Among them, watermark by hybrid halftone dots is formed by amplitude modulation (AM) dots and frequency modulation (FM) dots. Because of different halftone dots, watermark can be observed by closer look. In this study, we design a boundary processing to make the edge area of two dots gradual change. Then, the visibility is evaluated by Human Visual System (HVS). Results shows the boundary processing can effectively hide watermark, and it can be applied to many document security applications.

Holograms are the optical devices to reconstruct images by recovering amplitude and phase of light, which show many applications in our daily life. Recently, the metasurfaces, an array of sub-wavelength antenna, show the abilities to manipulate both the amplitude and phase of incident electromagnetic wave in a wide electromagnetic region from visible to microwave frequencies. Here, we realized a high-efficiency broadband plasmonic metasurface and metahologram by reflected metasurface made of subwavelength gold nanoantennas which is designed and fabricated in optical frequencies. The reconstructed images of our metahologram show polarization-controlled dual images with high contrast, functioning for both coherent and incoherent light sources within a broad spectral range and under a wide range of incidence angles. Aluminum plasmonics based metaholograms with multi-images, multi-channels, or multiplexing at different wavelengths, on the other hand, prefer narrower plasmonic resonances and less phase levels to realize multi-color metahologram. By combining with the techniques of tunable metasurfaces, metahologram can potentially be used to realize active hologram that works at arbitrary electromagnetic wave region.

Doped single crystalline fibers are efficient and broadband gain media for lasers, as well as cw and high-brightness emissions. Various crystal fibers based broadband tunable laser and emissions centered from blue to near infrared will be addressed. A Cr4+:YAG crystal fiber laser tunable from 1353 to 1509 nm with a threshold pump power of 70 mW will be discussed as an example.

A Rough Set (RS) based dataset reduction method using SWARM optimization algorithm and a cluster validation function is proposed. In the proposed approach, the user specifies the classification quality required in advance, and the method then finds the attribute reducts and perform attribute discretization to satisfy the desired quality of classification. While many other solutions are possible, the proposed method yields the solution which satisfies the optimal discretization conditions by means of a newly-designed cluster validation index function. The performance of the proposed method is compared with that of two existing attribute reduction and classification methods for eight benchmark datasets. The results confirm that the proposed method provides an effective tool for solving simultaneous attribute reduction and discretization problems.

Dictionary learning has been applied to computer vision problems such as facial expression recognition. K-SVD is one of the state-of-the-art dictionary learning algorithms. However, K-SVD is unsupervised and focuses only on the representational power. In this paper, we adopt label-consistent K-SVD with scattering transform in facial expression recognition. In addition to reducing the reconstruction error, label-consistent K-SVD combines further the discriminative sparse-code error and classification error in the optimization. Experimental results show that our approach can improve the performance of facial expression recognition when sparse coding is used.

This paper presents a driver assistance system to monitor the driver driving behavior by applying the so-called “First-Person Vision” (FPV) technology. It consists of two modules: the scene classification and the driver viewing angle estimation. First, we use “bag of words” image classification approach based on FAST and BRIEF feature descriptor in the dataset. Second, we establish the “vocabulary dictionary” to encode an input image as a feature vector. Third, we apply SVM classifier to detect whether the driver's view is inside or outside scene of a vehicle. Finally, we estimate the driver viewing angle estimation based on FPV and the windshield-mounted camera. In the experiments, we illustrate the effectiveness of our system.

In this paper, we propose a framework that estimates the discrete intensity rank of a facial expression based on a single image. For most people, judging whether an expression is more intense than others is easier than determining its real-valued intensity degree, and hence the relative order of two expressions is more distinguishable than the exact difference between them. We utilize the relative order to construct an image-based ranking approach for inferring the discrete ranks. The challenge of image-based approaches is to conduct a representation for subtle expression changes. We employ an efficient descriptor, scattering transform, which is translation invariant and can linearize deformations. This scattering representation recovers the lost high frequencies and retains discrimination under invariant property. Our experimental results demonstrate that the proposed framework with scattering transform outperforms other compared feature descriptors and algorithms.

We investigate the resonantly side-pumped 1645-nm Er:YAG crystal fiber laser utilizing a gold-embedded grating coupler. The 91.5% pump coupling efficiency and 73% laser slope efficiency can be achieved in a 18-cm double-clad crystal fiber laser.

The crystalline YAG-core fiber cladded by high-index N-LaSF9 glass has been fabricated to significantly reduce the number of guided modes. The selective propagation of the LP01 and LP11 modes was successfully demonstrated at 633 nm.

A function is proposed for descritizing and classifying the uncertain data of multi-attribute decision-making (MADM) datasets using a hybrid scheme incorporating fuzzy set theory, Rough Set (RS) theory and a modified form of the PBMF index function. The proposed MADM index function is used to extend the applicability of the single-attribute decision-making (SADM) function. The validity of the proposed MADM index function is evaluated by comparing the descritizing results obtained for a simple hypothetical function with those obtained using a SADM function and the conventional PBMF function.

With the rapid advancement of hardware efficiency in recent years, computation power has significantly improved, increasing the speed and depths of the game tree search in computer chess programs. However, to satisfy the need of increasing search depth, the investment in hardware is no longer personally affordable. Thus, another method must be found in addition to improving the efficiency of hardware and heuristic algorithms. In this paper, knowledge-based technologies are applied to design a database for the opening phase of MiniShogi. By analyzing game records and ranking the moves in advance, the program can increase search depths and reduce search time spent in the opening phase.

This study proposes a method for partitioning and classifying complex datasets using a hybrid method based on Fuzzy C-Means (FCM) method, Variable Precision Rough Set (VPRS) theory and a modified form of the PBMF index function (a cluster validity index function). The proposed VPRS index method partitions the attributes within the dataset rather than the data and achieves both the optimal number of clusters and the optimal classification accuracy. The validity of the proposed approach is confirmed by comparing the clustering results obtained from the VPRS method for a hypothetical function and a typical stock market system with those obtained from the conventional RS and PBMF methods, respectively. Overall, the results show that the VPRS index method not only has a better clustering performance than the PBMF method, but also achieves greater classification accuracy, and therefore provides a more reliable basis for the extraction of decision-making rules.

In this paper, we propose an ordinal hyperplane ranking algorithm called OHRank, which estimates human ages via facial images. The design of the algorithm is based on the relative order information among the age labels in a database. Each ordinal hyperplane separates all the facial images into two groups according to the relative order, and a cost-sensitive property is exploited to find better hyperplanes based on the classification costs. Human ages are inferred by aggregating a set of preferences from the ordinal hyperplanes with their cost sensitivities. Our experimental results demonstrate that the proposed approach outperforms conventional multiclass-based and regression-based approaches as well as recently developed ranking-based age estimation approaches.

The advantage of using channel feedback information to determine the optimal pilot placement is examined for amplify-and-forward OFDM relay links. This work extends upon the previously proposed feedback-aided pilot placement scheme to systems that employ subcarrier pairing at the relay. With channel feedback, the pilot subcarriers can be properly chosen to reduce the channel estimation error and increase the effective signal-to-noise ratio (SNR) of data subcarriers. When subcarrier pairing is employed, the relay reselects the subcarriers used to forward the source's data based on the effective SNR of each subcarrier, which in order is affected by the pilot placement policy. Here, the minimum effective SNR observed by all subcarriers' data is utilized as the performance measure since it is what dominates the symbol-error-rate (SER) at high SNR. For a given pilot placement, the best subcarrier pairing scheme is found to be the inverse mapping of the source-to-relay (SR) and relay-to-destination (RD) channels based on the quality of their effective SNR. A lower bound is derived for the gain achievable with any feedback-aided pilot placement scheme. Due to the complexity of finding the optimal pilot placement, a more efficient iterative pilot relocation (IPR) scheme is then proposed. In this scheme, the subcarrier with the worst equivalent channel is used to replace one of the originally selected pilot locations in each iteration. The efficacy of the proposed policy is demonstrated through numerical simulations.

High brightness diode pump source is becoming a matter of concern for developing high power fiber laser [1]. It is crucial to convert the high power emission but low brightness laser diode array source into a kind of high brightness source and maintain the same power level. Complex beam-shaping methods for each single emitter was used in the past [2,3]. In order to increase the efficiency and compactness, grating coupling mechanism is a promising way to achieve the high brightness pumping. In this paper, we demonstrate a high efficiency and high power fiber laser through the diode array pump by using grating coupler.

In this study, the Grey Relational Analysis (GRA) model is combined with Fuzzy C-Means (FCM) clustering scheme and Rough Set (RS) theory to create an automatic portfolio selection mechanism. In the proposed approach, 53 financial indices are collected automatically for each stock item every quarter and a GRA model is used to consolidate these indices into six predetermined financial ratios (Grey Relational Grades (GRGs)). The GRGs of the stock items are then clustered using a FCM scheme and the resulting cluster indices are processed using RS theory to identify the lower approximate set within the stock system. The stock items within the lower approximate set are filtered in accordance with established investment principles and the six GRGs of each surviving stock item are then consolidated to a single GRG indicating the overall merit of the corresponding stock item in terms of its ability to maximize the rate of return on the investment portfolio. It is shown that the rate of return on the investment portfolio selected using the proposed GRA/FCM/RS system is higher than the average rate of return predicted by the variation in the Taiwan Stock Exchange Capitalization Weighted Stock Index (TAIEX) over the same period.

This paper examines the use of channel feedback to determine the optimal pilot placement for channel estimation of amplify-and-forward OFDM relay links. With channel information from the previous symbol period, new pilot locations can be determined so that, on the one hand, channel estimation errors are controlled at a low level and, on the other hand, reliable subcarrier channels are made available for data transmission. The minimum effective SNR across subcarriers is a factor that dominates the symbol-error-rate (SER) at high SNR and, thus, is utilized as the performance measure in this work. First, a lower bound is derived for the achievable SNR gain, which is defined as the ratio between the effective SNR achieved with feedback and that without feedback (i.e., the equal-spaced pilot-placement scheme). The analytic results show that, with a fixed number of subcarriers, the feedback gain will grow as the number of pilot tones increases. Due to the complexity required in finding the optimal solution, a suboptimal but efficient algorithm, called the iterative pilot relocation (IPR) scheme, is proposed, where iterative exchanges between old pilot locations and the location of the worst data subcarrier are done to improve the minimum effective SNR. The efficacy of the proposed IPR scheme is demonstrated through numerical simulations.

Subcarrier allocation, power allocation, and partner selection algorithms are examined for amplify-and-forward cooperative multicarrier systems. Consider a network that consists of multiple cooperative pairs employing a two-phase cooperation scheme where the users first transmit their own messages on each of their subcarriers in Phase I and either retransmit their own messages or relay the messages of their partners in Phase II, i.e., the cooperative phase. In general, subcarrier allocation and partner selection problems are known to be intractable for systems with large number of users and subcarriers. In this work, we propose a mathematically tractable approach to address these issues and show its effectiveness compared to existing algorithms. Specifically, we consider a relaxed problem formulation, where each subcarrier is able to retransmit or relay for all other subcarriers in Phase II over orthogonal channels, and derive the optimal power allocation for this scenario. Efficient subcarrier allocation policies can then be derived based on the power allocation result and a sufficient condition is obtained to determine when cooperation is helpful among two users. A partner selection algorithm is then devised based on the sufficient condition. Given the subcarrier allocation scheme, the optimal power allocation among subcarriers in the cooperative phase can be shown to be a convex optimization problem and, thus, can be solved efficiently using standard convex optimization tools. The efficacy of the proposed algorithms is demonstrated through numerical simulations.

In our daily life, it is much easier to distinguish which person is elder between two persons than how old a person is. When inferring a person's age, we may compare his or her face with many people whose ages are known, resulting in a series of comparative results, and then we conjecture the age based on the comparisons. This process involves numerous pairwise preferences information obtained by a series of queries, where each query compares the target person's face to those faces in a database. In this paper, we propose a ranking-based framework consisting of a set of binary queries. Each query collects a binary-classification-based comparison result. All the query results are then fused to predict the age. Experimental results show that our approach performs better than traditional multi-class-based and regression-based approaches for age estimation.

This study investigates the effects of toluene adsorption on the filtration efficiency of nanoparticles by a fluidized-bed adsorption reactor. Simulated flue gas is modified by different types of nanoparticle (SiO 2 and Al 2 O 3 ) with various particle sizes (10 and 80 nm) are also carried out in experiment. Therefore, the experimental results show that the removal efficiencies of the nanoparticles and toluene are in the range of 97% to 87% and 99.8% to 99.4%, respectively. Particularly, BET results show that the particles of 10 nm SiO 2 increase obstruction of the adsorbent volume with increasing toluene content. This finding shows that filtration efficiency increases with the addition of toluene.