Many selected mapping (SLM) schemes have been proposed to reduce the peak-to-average power ratio (PAPR) of orthogonal frequency division multiplexing (OFDM) signal sequences. In this paper, an efficient selection (ES) method of the OFDM signal sequence with minimum PAPR among many alternative OFDM signal sequences is proposed; it supports various SLM schemes. Utilizing the fact that OFDM signal components can be sequentially generated in many SLM schemes, the generation and PAPR observation of the OFDM signal sequence are processed concurrently. While the u-th alternative OFDM signal components are being generated, by applying the proposed ES method, the generation of that alternative OFDM signal components can be interrupted (or stopped) according to the selection criteria of the best OFDM signal sequence in the considered SLM scheme. Such interruption substantially reduces the average computational complexity of SLM schemes without degradation of PAPR reduction performance, which is confirmed by analytical and numerical results. Note that the proposed method is not an isolated SLM scheme but a subsidiary method which can be easily adopted in many SLM schemes in order to further reduce the computational complexity of considered SLM schemes.

An ordered successive interference cancellation (OSIC) scheme based on combined post-processing signal-to-interference-plus-noise ratio (PSINR) is proposed for multiple-input multiple-output (MIMO) systems with retransmission. For the OSIC procedures at the current transmission round, instead of reusing the PSINRs and decision statistics calculated for the previous transmission rounds, the proposed OSIC scheme newly calculates the combined PSINRs and combined decision statistics from the available receive signal vectors and channel matrices at every retransmission. Therefore, the proposed OSIC scheme utilizes all receive signal vectors and channel matrices obtained up to the current transmission round during the OSIC procedures. A low-complexity version of the proposed OSIC scheme is also proposed, and the low-complexity version recalculates the combined PSINRs and combined decision statistics from part of the available receive signal vectors and channel matrices. Simulation results verify that the proposed schemes achieve significantly better error performance than existing OSIC schemes based on the detection and combining process for MIMO systems with retransmission.

In this study, we propose a simple, yet general and powerful framework of integrating multiple global and local features by Product Sparse Coding (PSC) for image retrieval. In our framework, multiple global and local features are extracted from images and then are transformed to Trimmed-Root (TR)-features. After that, the features are encoded into compact codes by PSC. Finally, a two-stage ranking strategy is proposed for indexing in retrieval. We make three major contributions in this study. First, we propose TR representation of multiple image features and show that the TR representation offers better performance than the original features. Second, the integrated features by PSC is very compact and effective with lower complexity than by the standard sparse coding. Finally, the two-stage ranking strategy can balance the efficiency and memory usage in storage. Experiments demonstrate that our compact image representation is superior to the state-of-the-art alternatives for large-scale image retrieval.

This paper proposes a moving source localization method that combines TDOA, FDOA and doppler rate measurements. First, the observation equations are linearized by introducing nuisance variables and an initial solution of all the variables is acquired using the weighted least squares method. Then, the Taylor expression and gradient method is applied to eliminate the correlation between the elements in the initial solution and obtain the final estimation of the source position and velocity. The proposed method achieves CRLB derived using TDOA, FDOA and doppler rate and is much more accurate than the conventional TDOA/FDOA based method. In addition, it can avoid the rank-deficiency problem and is more robust than the conventional method. Simulations are conducted to examine the algorithm's performance and compare it with conventional TDOA/FDOA based method.

Adding versatile interactions to imperative programming - C, Java and Android - is an essential task. Unfortunately, existing languages provide only limited constructs for user interaction. These constructs are usually in the form of unbounded quantification. For example, existing languages can take the keyboard input from the user only via the read(x)/scan(x) statement. Note that the value of x is unbounded in the sense that x can have any value. This statement is thus not useful for applications with bounded inputs. To support bounded choices, we propose new bounded-choice statements for user interation. Each input device (keyboard, mouse, touchpad, ...) naturally requires a new bounded-choice statement. To make things simple, however, we focus on a bounded-choice statement for keyboard - kchoose - to allow for more controlled and more guided participation from the user. We illustrate our idea via CBI, an extension of the core C with a new bounded-choice statement for the keyboard.

In this paper, we propose a method to estimate the most resource-consuming disease from electronic claim data based on Labeled Latent Dirichlet Allocation (Labeled LDA). The proposed method models each electronic claim from its medical procedures as a mixture of resource-consuming diseases. Thus, the most resource-consuming disease can be automatically estimated by applying Labeled LDA to the electronic claim data. Although our method is composed of a simple scheme, this is the first trial for realizing estimation of the most resource-consuming disease.

Low pin-count testing is an effective method to reduce test cost. Based on this method multi-site testing, i.e., where multiple devices are tested concurrently, can be supported under the limitation on the number of channels provided by ATE. In this work we propose a scalable test module (called STM) design that can support multi-site testing more efficiently when compared with previous work. In the previous work, the total number of devices that can be tested concurrently is usually fixed when the design for testability hardware is designed. For our STM, each STM can deal with a number of circuits to be tested at the same time. Moreover, STM is scalable, i.e., multiple STMs can work collaboratively while the ATE bandwidth still remains the same to further increase the degree of test parallelism. Our STM will be integrated with ATE and serve as an interface between ATE and circuits under test (CUT). Only four pins are required by STM to communicate with ATE, and IEEE 1149.1 Std. ports are employed to transfer test data to/from CUTs. STM has been verified via silicon proof, which contains only about 2,768 logic gates. Experiments results for a number of ISCAS and IWLS'05 benchmark circuits also demonstrate that by making good use of the scalable feature of STM, test efficiency can be enhanced significantly.

In the field of education such as elementary and middle schools, teachers want to take care of schoolchildren during physical trainings and after-school club activities. On the other hand, in the field of sports such as professional and national-level sports, physical or technical trainers want to manage the health, physical and physiological conditions of athletes during exercise trainings in the grounds. In this way, it is required to monitor vital signs for persons during exercises, however, there are several technical problems to be solved in its realization. In this paper, we present the importance and necessity of vital monitoring for persons during exercises, and to make it possible periodically, reliably and in real-time, we present the solutions which we have so far worked out and point out remaining technical challenges in terms of vital/physical sensing, wireless transmission and human interface.

Change-aware development environments can automatically record fine-grained code changes on a program and allow programmers to replay the recorded changes in chronological order. However, since they do not always need to replay all the code changes to investigate how a particular entity of the program has been changed, they often eliminate several code changes of no interest by manually skipping them in replaying. This skipping action is an obstacle that makes many programmers hesitate when they use existing replaying tools. This paper proposes a slicing mechanism that automatically removes manually skipped code changes from the whole history of past code changes and extracts only those necessary to build a particular class member of a Java program. In this mechanism, fine-grained code changes are represented by edit operations recorded on the source code of a program and dependencies among edit operations are formalized. The paper also presents a running tool that slices the operation history and replays its resulting slices. With this tool, programmers can avoid replaying nonessential edit operations for the construction of class members that they want to understand. Experimental results show that the tool offered improvements over conventional replaying tools with respect to the reduction of the number of edit operations needed to be examined and over history filtering tools with respect to the accuracy of edit operations to be replayed.

Efficient detection schemes for an overloaded multiple-input multiple-output (MIMO) system have been investigated recently. The literature shows that trellis coded modulation (TCM) is able to enhance a system's capability to separate signal streams in the detection process of MIMO systems. However, the computational complexity remains high as a maximum likelihood detection (MLD) algorithm is used in the scheme. Thus, a sphere decoding (SD) algorithm with a pseudo distance (PD) is proposed in this paper. The PD maintains the coding gain advantage of the TCM by keeping some potential paths connected unlike conventional SD which truncates them. It is shown that the proposed scheme can reduce the number of distance calculations by about 98% for the transmission of 3 signal streams. In addition, the proposed scheme improves the performance by about 2dB at the bit error rate of 10-2.

We numerically and experimentally demonstrate for the first time a novel all-optical quantization technique using dense spectral slicing with a specially designed arrayed waveguide grating for orthogonal frequency division multiplexed signals. By using a mode-locked laser diode with low jitter, the quantization technique can be achieved a high-speed and low-jitter operation. Both numerical and experimental results confirm the feasibility of 10 GSample/s, completely linear 3-bit step quantization for photonic analog to digital conversion. This optical quantization technique will be beneficial for ultra-high-speed optical communication using digital signal processing.

Research into closed-form Gaussian sum smoother has provided an attractive approach for tracking in clutter, joint detection and tracking (in clutter), and multiple target tracking (in clutter) via the probability hypothesis density (PHD). However, Gaussian sum smoother with nonlinear target model has particular nonlinear expressions in the backward smoothed density that are different from the other filters and smoothers. In order to extend the closed-form solution of linear Gaussian sum smoother to nonlinear model, two closed-form approximations for nonlinear Gaussian sum smoother are proposed, which use Gaussian particle approximation and unscented transformation approximation, separately. Since the estimated target number of PHD smoother is not stable, a heuristic approximation method is added. At last, the Bernoulli smoother and PHD smoother are simulated using Gaussian particle approximation and unscented transformation approximation, and simulation results show that the two proposed algorithms can obtain smoothed tracks with nonlinear models, and have better performance than filter.

Out-of-order superscalar processors rename register numbers to remove false dependencies between instructions. A renaming logic for register renaming is a high-cost module in a superscalar processor, and it consumes considerable energy. A renamed trace cache (RTC) was proposed for reducing the energy consumption of a renaming logic. An RTC caches and reuses renamed operands, and thus, register renaming can be omitted on RTC hits. However, conventional RTCs suffer from several performance, energy consumption, and hardware overhead problems. We propose a semi-global renamed trace cache (SGRTC) that caches only renamed operands that are short distance from producers outside traces, and solves the problems of conventional RTCs. Evaluation results show that SGRTC achieves 64% lower energy consumption for renaming with a 0.2% performance overhead as compared to a conventional processor.

We design a new hash table for high-speed networking that reduces main memory accesses even when the ratio of inserted items to the table size is high, at which point previous schemes no longer work. This improvement comes from a new design of a summary, called expanded keys, exploiting recent multiple hash functions and Bloom filter theories.

In this study, we propose a method for measuring a photoplethysmograph using a complementary metal-oxide-semiconductor image sensor (CMOS) or smartphone camera for the adaptation of a mobile health (m-health) services. The proposed algorithm consists of six procedures. Before measuring the photoplethysmograph, the human fingertip must make contact with the smartphone camera lens and turn on the camera light. The first procedure converts the red-green-blue (RGB) to a gray image from a camera image, Then, region of interest (ROI) must be detected from the obtained image. The third procedure calculates the baseline level to reduce direct current (DC) offset effect, before extracting the photoplethysmograph from the camera image. The baseline is filtered, and the last step oversamples the resulting baseline filtered data using cubic spline interpolation. The proposed algorithm has been tested on six people using CMOS image sensors of several smartphones, which can effectively acquire a PPG signal in any situation. We believe that the proposed algorithm could easily be adapted into any m-health system that used a CMOS image sensor.

The the depth-of-field limitation of our eyes causes out-of-focus blur in the retinal images. The blur dynamically changes whenever we change our gaze and accordingly the scene point we are looking at changes its depth. This paper proposes an image display that reproduces retinal out-of-focus blur by using a stereoscopic display and eye trackers. Its purpose is to provide the viewer with more realistic visual experiences than conventional (stereoscopic) displays. Unlike previous similar systems that track only one of the viewer's eyes to estimate the gaze depth, the proposed system tracks both eyes individually using two eye trackers and estimates the gaze depth from the convergence angle calculated by triangulation. This provides several advantages over existing schemes, such as being able to deal with scenes having multiple depths. We describe detailed implementations of the proposed system and show the results of an experiment conducted to examine its effectiveness. In the experiment, creating a scene having two depths using two LCD displays together with a half mirror, we examined how difficult it is for viewers to distinguish between the real scene and its virtual reproduction created by the proposed display system. The results of the experiment show the effectiveness of the proposed approach.

I report modulation format conversion technology that maps on-off keying to 4-level pulse amplitude modulation signals. The conversion technology is based on balanced detection and intensity modulation. Two input optical on-off keying signals into a balanced photo detector produce an electrical signal that drives an intensity modulator to generate an optical pulse amplitude modulation signal. Two 20Gbit/s on-off keying signals were successfully converted into a 40Gbit/s pulse amplitude modulation signal.

We consider a problem as follows: Given unit weights arriving in an online manner with the total cardinality unknown, upon each arrival we decide where to place it on the unit circle in $mathbb{R}^{2}$. The objective is to set the center of mass of the placed weights as close to the origin as possible. We apply competitive analysis defining the competitive difference as a performance measure. We first present an optimal strategy for placing unit weights which achieves a competitive difference of $rac{1}{5}$. We next consider a variant in which the destination of each weight must be chosen from a set of positions that equally divide the unit circle. We give a simple strategy whose competitive difference is 0.35. Moreover, in the offline setting, several conditions for the center of mass to lie at the origin are derived.

In this paper we propose an adaptive bitrate (ABR) algorithm that selects the video rates by observing and controlling the playback buffer. In a Hypertext Transfer Protocol (HTTP) adaptive streaming session, buffer dynamics largely depend on the chunk sizes. First, we present the algorithm that selects the next video rates based on the current buffer level, while considering the upcoming chunk sizes. We aim to exploit the variation of chunk sizes of a variable bitrate (VBR) encoded video to optimize the tradeoff between the video rate and buffer underflow events while keeping a low frequency of video rate changes. To evaluate the performance of the proposed algorithm, we consider three scenarios: (i) the video flow does not compete with any cross traffic, (ii) the video flow shares the bottleneck link with competing TCP traffic, and (iii) two HTTP clients share the bottleneck. We show that the proposed algorithm selects a high playback video rate and avoids unnecessary rebuffering while keeping a low frequency of video rate changes. Furthermore, we show that the proposed algorithm changes the video quality gradually to guarantee the user's viewing experience.

Many methods have been proposed for measuring the structural similarity between two protein folds. However, it is difficult to select one best method from them for the classification task, as each method has its own strength and weakness. Intuitively, combining multiple methods is one solution to get the optimal classification results. In this paper, by generalizing the concept of the large margin nearest neighbor (LMNN), a method for combining multiple distance metrics from different types of protein structure comparison methods for protein fold classification task is proposed. While LMNN is limited to Mahalanobis-based distance metric learning from a set of feature vectors of training data, the proposed method learns an optimal combination of metrics from a set of distance metrics by minimizing the distances between intra-class data and enlarging the distances of different classes' data. The main advantage of the proposed method is the capability in finding an optimal weight coefficient for combination of many metrics, possibly including poor metrics, avoiding the difficulties in selecting which metrics to be included for the combination. The effectiveness of the proposed method is demonstrated on classification experiments using two public protein datasets, namely, Ding Dubchak dataset and ENZYMES dataset.