Realization of a planar dual-band fork three-way power divider (PDBF3PD) with Cheng's equivalent structure is discussed. The Cheng's structure consists of two open-circuited stubs and a transmission line, and the characteristic impedances tend to be high. As a result, the realizable range of frequency ratios of upper frequency to lower frequency is limited in a narrow area. In this paper, an impedance scale factor is proposed to transform characteristic impedances into a realizable range and to facilitate the design of PDBF3PDs. Theoretical considerations are verified using a simulator of ADS2008U and by an experiment.

Local descriptors, Local Binary Pattern (LBP) and Scale Invariant Feature Transform (SIFT) are widely used in various computer applications. They emphasize different aspects of image contents. In this letter, we propose to combine them in sparse coding for categorizing scene images. First, we regularly extract LBP and SIFT features from training images. Then, corresponding to each feature, a visual word codebook is constructed. The obtained LBP and SIFT codebooks are used to create a two-dimensional table, in which each entry corresponds to an LBP visual word and a SIFT visual word. Given an input image, LBP and SIFT features extracted from the same positions of this image are encoded together based on sparse coding. After that, spatial max pooling is adopted to determine the image representation. Obtained image representations are converted into one-dimensional features and classified by utilizing SVM classifiers. Finally, we conduct extensive experiments on datasets of Scene Categories 8 and MIT 67 Indoor Scene to evaluate the proposed method. Obtained results demonstrate that combining features in the proposed manner is effective for scene categorization.

In this letter, a novel antenna selection (AS) technique is proposed for the downlink of large-scale multi-user multiple input multiple output (MU-MIMO) networks, where a base station (BS) is equipped with large-scale antennas (N) and communicates simultaneously with K(K ≪ N) mobile stations (MSs). In the proposed scheme, the S antennas (S ≤ N) are selected by utilizing the concept of a sliding window. It is shown that the sum-rate of our proposed scheme is comparable to that of the conventional scheme, while the proposed scheme can significantly reduce the complexity of the BS.

Smart city services are implemented using various data collected from houses and infrastructure within a city. As the volume and variety of the smart city data becomes huge, individual services have suffered from expensive computation effort and large processing time. In order to reduce the effort and time, this paper proposes a concept of Materialized View as a Service (MVaaS). Using the MVaaS, every application can easily and dynamically construct its own materialized view, in which the raw data is converted and stored in a convenient format with appropriate granularity. Thus, once the view is constructed, the application can quickly access necessary data. In this paper, we design a framework of MVaaS specifically for large-scale house log, managed in a smart-city data platform. In the framework, each application first specifies how the raw data should be filtered, grouped and aggregated. For a given data specification, MVaaS dynamically constructs a MapReduce batch program that converts the raw data into a desired view. The batch is then executed on Hadoop, and the resultant view is stored in HBase. We present case studies using house log in a real home network system. We also conduct an experimental evaluation to compare the response time between cases with and without MVaaS.

In this paper, we present a differential dual-modulus prescaler based on an injection-locked frequency divider (ILFD) for satellite low-noise block (LNB) down-converters. We fabricated three-stage differential latches using an ILFD and a cascaded differential divider in a 130-nm CMOS process. The prototype chip core area occupies 40µm × 20µm. The proposed prescaler achieved the locking range of 2.1-10GHz with both divide-by-10 and divide-by-11 operations at a supply voltage of 1.4V. Normalized energy consumptions are 0.4pJ (=mW/GHz) at a 1.4-V supply voltage and 0.24pJ at a 1.2-V supply voltage. To evaluate the tolerance of phase-difference deviation of the input differential pair from the perfect differential phase-difference, 180 degrees, we measured the operational frequencies for various phase-difference inputs. The proposed prescaler achieved the operational frequency range of 2.1-10GHz with an input phase-difference deviation of less than 90 degrees. However, the range of operational frequency decreases as the phase-difference deviation increases beyond 90 degrees and reaches 3.9-7.9GHz for the phase-difference deviation of 180 degrees (i.e. no phase difference). In addition, to confirm the fully locking operation, we measured the spurious noise and the phase noise degradation while reducing the supply voltage. The sensitivity analysis of the prescaler for various supply voltages can explain the above degradation of spectral purity. Spurious noise arises and the phase noise degrades with decreasing supply voltage due to the quasi- and non-locking operations. We verified the fully-locking operation for the LNB down-converter at a 1.4-V supply voltage.

This letter presents a new entropy measure for electroencephalograms (EEGs), which reflects the underlying dynamics of EEG over multiple time scales. The motivation behind this study is that neurological signals such as EEG possess distinct dynamics over different spectral modes. To deal with the nonlinear and nonstationary nature of EEG, the recently developed empirical mode decomposition (EMD) is incorporated, allowing an EEG to be decomposed into its inherent spectral components, referred to as intrinsic mode functions (IMFs). By calculating Shannon entropy of IMFs in a time-dependent manner and summing them over adaptive multiple scales, the result is an adaptive subscale entropy measure of EEG. Simulation and experimental results show that the proposed entropy properly reveals the dynamical changes over multiple scales.

We describe a large-scale integrated circuit (LSI) design of rapid single-flux-quantum (RSFQ) circuits and demonstrate several reconfigurable data-path (RDP) processor prototypes based on the ISTEC Advanced Process (ADP2). The ADP2 LSIs are made up of nine Nb layers and Nb/AlOx/Nb Josephson junctions with a critical current density of 10kA/cm2, allowing higher operating frequencies and integration. To realize truly large-scale RSFQ circuits, careful design is necessary, with several compromises in the device structure, logic gates, and interconnects, balancing the competing demands of integration density, design flexibility, and fabrication yield. We summarize numerical and experimental results related to the development of a cell-based design in the ADP2, which features a unit cell size reduced to 30-µm square and up to four strip line tracks in the unit cell underneath the logic gates. The ADP LSIs can achieve ∼10 times the device density and double the operating frequency with the same power consumption per junction as conventional LSIs fabricated using the Nb four-layer process. We report the design and test results of RDP processor prototypes using the ADP2 cell library. The RDP processors are composed of many arrays of floating-point units (FPUs) and switch networks, and serve as accelerators in a high-performance computing system. The prototypes are composed of two-dimensional arrays of several arithmetic logic units instead of FPUs. The experimental results include a successful demonstration of full operation and reconfiguration in a 2×2 RDP prototype made up of 11.5k junctions at 45GHz after precise timing design. Partial operation of a 4×4 RDP prototype made up of 28.5k-junctions is also demonstrated, indicating the scalability of our timing design.

Packet loss and energy dissipation are two major challenges of designing large-scale wireless sensor networks. Since sensing data is spatially correlated, compressed sensing (CS) is a promising reconstruction scheme to provide low-cost packet error correction and load balancing. In this letter, assuming a multi-hop network topology, we present a CS-oriented data aggregation scheme with a new measurement matrix which balances energy consumption of the nodes and allows for recovery of lost packets at fusion center without additional transmissions. Comparisons with existing methods show that the proposed scheme offers higher recovery precision and less energy consumption on TinyOS.

In this paper, an efficient method to reduce computational complexity for pedestrian detection is presented. Since trilinear interpolation is not used, the amount of required operations for histogram of oriented gradient (HOG) feature calculation is significantly reduced. By calculating multi-scale HOG features with integral HOG in a two-stage approach, both high detection rate and speed are achieved in the proposed method.

Checkpoint-rollback recovery, which is a universal method for restoring distributed systems after faults, requires a sophisticated snapshot algorithm especially if the systems are large-scale, since repeatedly taking global snapshots of the whole system requires unacceptable communication cost. As a sophisticated snapshot algorithm, a partial snapshot algorithm has been introduced that takes a snapshot of a subsystem consisting only of the nodes that are communication-related to the initiator instead of a global snapshot of the whole system. In this paper, we modify the previous partial snapshot algorithm to create a new one that can take a partial snapshot more efficiently, especially when multiple nodes concurrently initiate the algorithm. Experiments show that the proposed algorithm greatly reduces the amount of communication needed for taking partial snapshots.

We present an algorithm for simultaneously recognizing and localizing planar textured objects in an image. The algorithm can scale efficiently with respect to a large number of objects added into the database. In contrast to the current state-of-the-art on large scale image search, our algorithm can accurately work with query images consisting of several specific objects and/or multiple instances of the same object. Our proposed algorithm consists of two major steps. The first step is to generate a set of hypotheses that provides information about the identities and the locations of objects in the image. To serve this purpose, we extend Bag-Of-Visual-Word (BOVW) image retrieval by incorporating a re-ranking scheme based on the Hough voting technique. Subsequently, in the second step, we propose a geometric verification algorithm based on A Contrario decision framework to draw out the final detection results from the generated hypotheses. We demonstrate the performance of the algorithm on the scenario of recognizing CD covers with a database consisting of more than ten thousand images of different CD covers. Our algorithm yield to the detection results of more than 90% precision and recall within a few seconds of processing time per image.

The link structure of the Web is generally viewed as a webgraph. One of the main objectives of web structure mining is to find hidden communities on the Web based on the webgraph, and one of its approaches tries to enumerate substructures, each of which corresponds to a set of web pages of a community or its core. Research has shown that certain substructures can find sets of pages that are inherently irrelevant to communities. In this paper, we propose a model, which we call contracted webgraphs, where such substructures are contracted into single nodes to hide useless information. We then try structure mining iteratively on those contracted webgraphs since we can expect to find further hidden information once irrelevant information is eliminated. We also explore the structural properties of contracted webgraphs from the viewpoint of scale-freeness, and we observe that they exhibit novel and extreme self-similarities.

As a network evolves following initial deployment, its service functions remain diversified through the openness of the network functions. This indicates that appropriate simplification of the service functions is essential if the evolving network is to achieve the required scalability of service processing and service management. While the screening of service functions is basically performed by network users and the market, several service functions will be automatically simplified based on the growth of the evolving network. This paper verifies the simplification of service functions resulting from the evolution of the network itself. First, the principles that serve as the basis for simplifying the service functions are explained using several practical examples. Next, a simulation model is proposed to verify the simplification of service functions in terms of the priority control function for path routing and load balancing among multiple paths. From the results of the simulation, this study clarifies that the anticipated simplification of service functions is actually realizable and the service performance requirements can be reduced as the network evolves after deployment. When the simplification of service functions can improve network quality, it accelerates the evolution of the network and increases the operator's revenue.

This letter introduces innovative VAD based on horizontal spectral entropy with long-span of time (HSELT) feature sets to improve mobile ASR performance in low signal-to-noise ratio (SNR) conditions. Since the signal characteristics of nonstationary noise change with time, we need long-term information of the noisy speech signal to define a more robust decision rule yielding high accuracy. We find that HSELT measures can horizontally enhance the transition between speech and non-speech segments. Based on this finding, we use the HSELT measures to achieve high accuracy for detecting speech signal form various stationary and nonstationary noises.

Face retrieval in news video has been identified as a challenging task due to the huge variations in the visual appearance of the human face. Although several approaches have been proposed to deal with this problem, their extremely high computational cost limits their scalability to large-scale video datasets that may contain millions of faces of hundreds of characters. In this paper, we introduce approaches for face retrieval that are scalable to such datasets while maintaining competitive performances with state-of-the-art approaches. To utilize the variability of face appearances in video, we use a set of face images called face-track to represent the appearance of a character in a video shot. Our first proposal is an approach for extracting face-tracks. We use a point tracker to explore the connections between detected faces belonging to the same character and then group them into one face-track. We present techniques to make the approach robust against common problems caused by flash lights, partial occlusions, and scattered appearances of characters in news videos. In the second proposal, we introduce an efficient approach to match face-tracks for retrieval. Instead of using all the faces in the face-tracks to compute their similarity, our approach obtains a representative face for each face-track. The representative face is computed from faces that are sampled from the original face-track. As a result, we significantly reduce the computational cost of face-track matching while taking into account the variability of faces in face-tracks to achieve high matching accuracy. Experiments are conducted on two face-track datasets extracted from real-world news videos, of such scales that have never been considered in the literature. One dataset contains 1,497 face-tracks of 41 characters extracted from 370 hours of TRECVID videos. The other dataset provides 5,567 face-tracks of 111 characters observed from a television news program (NHK News 7) over 11 years. We make both datasets publically accessible by the research community. The experimental results show that our proposed approaches achieved a remarkable balance between accuracy and efficiency.

This paper presents a non-blind watermarking technique that is robust to non-linear geometric distortion attacks. This is one of the most challenging problems for copyright protection of digital content because it is difficult to estimate the distortion parameters for the embedded blocks. In our proposed scheme, the location of the blocks are recorded by the translation parameters from multiple Scale Invariant Feature Transform (SIFT) feature points. This method is based on two assumptions: SIFT features are robust to non-linear geometric distortion and even such non-linear distortion can be regarded as “linear” distortion in local regions. We conducted experiments using 149,800 images (7 standard images and 100 images downloaded from Flickr, 10 different messages, 10 different embedding block patterns, and 14 attacks). The results show that the watermark detection performance is drastically improved, while the baseline method can achieve only chance level accuracy.

This paper presents a novel scale-rotation invariant generative model (SRIGM) and a kernel sparse representation classification (KSRC) method for scene categorization. Recently the sparse representation classification (SRC) methods have been highly successful in a number of image processing tasks. Despite its popularity, the SRC framework lucks the abilities to handle multi-class data with high inter-class similarity or high intra-class variation. The kernel random coordinate descent (KRCD) algorithm is proposed for 1 minimization in the kernel space under the KSRC framework. It allows the proposed method to obtain satisfactory classification accuracy when inter-class similarity is high. The training samples are partitioned in multiple scales and rotated in different resolutions to create a generative model that is invariant to scale and rotation changes. This model enables the KSRC framework to overcome the high intra-class variation problem for scene categorization. The experimental results show the proposed method obtains more stable performances than other existing state-of-art scene categorization methods.

Existing large-scale systems suffer from various hardware/software failures, motivating the research of fault-tolerance techniques. Checkpoint-restart techniques are widely applied fault-tolerance approaches, especially in scientific computing systems. However, the overhead of checkpoint largely influences the overall system performance. Recently, the emerging byte-addressable, persistent memory technologies, such as phase change memory (PCM), make it possible to implement checkpointing in arbitrary data granularity. However, the impact of data granularity on the checkpointing cost has not been fully addressed. In this paper, we investigate how data granularity influences the performance of a checkpoint system. Further, we design and implement a high-performance checkpoint system named AG-ckpt. AG-ckpt is a hybrid-granularity incremental checkpointing scheme through: (1) low-cost modified-memory detection and (2) fine-grained memory duplication. Moreover, we also formulize the performance-granularity relationship of checkpointing systems through a mathematical model, and further obtain the optimum solutions. We conduct the experiments through several typical benchmarks to verify the performance gain of our design. Compared to conventional incremental checkpoint, our results show that AG-ckpt can reduce checkpoint data amount up to 50% and provide a speedup of 1.2x-1.3x on checkpoint efficiency.

Conventional entropy measure is derived from full-band (range from 0 Hz to 4 kHz); however, it can not clearly describe the spectrum variability during voice-activity. Here we propose a novel concept of adaptive long-term sub-band entropy ( ALT-SubEnpy ) measure and combine it with a multi-thresholding scheme for voice activity detection. In detail, the ALT-SubEnpy measure developed with four part parameters of sub-entropy which uses different long-term spectral window length at each part. Consequently, the proposed ALT-SubEnpy -based algorithm recursively updates the four adaptive thresholds on each part. The proposed ALT-SubEnpy-based VAD method is shown to be an effective method while working at variable noise-level condition.

In this paper, we present a new fault-tolerant, large-scale star network scheme called Scalable Autonomous Fault-tolerant Ethernet (SAFE). The primary goal of a SAFE scheme is to provide network scalability and autonomous fault detection and recovery. SAFE divides a large-scale, mission-critical network, such as the naval combatant network, into several subnets by limiting the number of nodes in each subnet. This network can be easily configured as a star network in order to meet fault recovery time requirements. For SAFE, we developed a novel mechanism for inter-subnet fault detection and recovery; a conventional Ethernet-based heartbeat mechanism is used in each subnet. Theoretical and experimental performance analyses of SAFE in terms of fail-over time were conducted under various network failure scenarios. The results validate our scheme.