System expandability becomes a major concern for highly parallel computers and data centers, because their number of nodes gradually increases year by year. In this context we propose a low-degree topology and its floor layout in which a cabinet or node set can be newly inserted by connecting short cables to a single existing cabinet. Our graph analysis shows that the proposed topology has low diameter, low average shortest path length and short average cable length comparable to existing topologies with the same degree. When incrementally adding nodes and cabinets to the proposed topology, its diameter and average shortest path length increase modestly. Our discrete-event simulation results show that the proposed topology provides a comparable performance to 2-D Torus for some parallel applications. The network cost and power consumption of DSN-F modestly increase when compared to the counterpart non-random topologies.

View full abstract

This paper presents the design of a multiple-standard 1080 high definition (HD) video decoder on a mixed-grained reconfigurable computing platform integrating coarse-grained reconfigurable processing units (RPUs) and FPGAs. The proposed RPU, including 16×16 multi-functional processing elements (PEs), is used to accelerate compute-intensive tasks in the video decoding. A soft-core-based microprocessor array is implemented on the FPGA and adopted to speed-up the dynamic reconfiguration of the RPU. Furthermore, a mail-box-based communication scheme is utilized to improve the communication efficiency between RPUs and FPGAs. By exploiting dynamic reconfiguration of the RPUs and static reconfiguration of the FPGAs, the proposed platform achieves scalable performances and cost trade-offs to support a variety of video coding standards, including MPEG-2, AVS, H.264, and HEVC. The measured results show that the proposed platform can support H.264 1080 HD video streams at up to 57 frames per second (fps) and HEVC 1080 HD video streams at up to 52fps under 250MHz, at the same time, it achieves a 3.6× performance gain over an industrial coarse-grained reconfigurable processor for H.264 decoding, and a 6.43× performance boosts over a general purpose processor based implementation for HEVC decoding.

View full abstract

Most Android applications are written in JAVA and run on a Dalvik virtual machine. For smartphone vendors and users who wish to know the performance of an application on a particular smartphone but cannot obtain the source code, we propose a new technique, Dalvik Profiler for Applications (DPA), to profile an Android application on a Dalvik virtual machine without the support of source code. Within a Dalvik virtual machine, we determine the entry and exit locations of a method, log its execution time, and analyze the log to determine the performance of the application. Our experimental results show an error ratio of less than 5% from the baseline tool Traceview which instruments source code. The results also show some interesting behaviors of applications and smartphones: the performance of some smartphones with higher hardware specifications is 1.5 times less than the phones with lower specifications. DPA is now publicly available as an open source tool.

View full abstract

GitHub is a developers' social networking service that hosts a great number of open source software (OSS) projects. Although some of the hosted projects are growing and have many developers, most projects are organized by a few developers and face difficulties in terms of sustainability. OSS projects depend mainly on volunteer developers, and attracting and retaining these volunteers are major concerns of the project stakeholders. To investigate the population structures of OSS development communities in detail and conduct software analytics to obtain actionable information, we apply a demographic approach. Demography is the scientific study of population and seeks to identify the levels and trends in the size and components of a population. This paper presents a case study, investigating the characteristics of the population structures of OSS projects on GitHub, and shows population projections generated with the well-known cohort component method. We found that there are four types of population structures in OSS development communities in terms of experiences and contributions. In addition, we projected the future population accurately using a cohort component population projection method. This method predicts a population of the next period using a survival rate calculated from past population. To the best of our knowledge, this is the first study that applied demography to the field of OSS research. Our approach addressing OSS-related problems based on demography will bring new insights, since studying population is novel in OSS research. Understanding current and future structures of OSS projects can help practitioners to monitor a project, gain awareness of what is happening, manage risks, and evaluate past decisions.

View full abstract

In this paper, we exploit MapReduce framework and other optimizations to improve the performance of hash join algorithms on multi-core CPUs, including No partition hash join and partition hash join. We first implement hash join algorithms with a shared-memory MapReduce model on multi-core CPUs, including partition phase, build phase, and probe phase. Then we design an improved cuckoo hash table for our hash join, which consists of a cuckoo hash table and a chained hash table. Based on our implementation, we also propose two optimizations, one for the usage of SIMD instructions, and the other for partition phase. Through experimental result and analysis, we finally find that the partition hash join often outperforms the No partition hash join, and our hash join algorithm is faster than previous work by an average of 30%.

View full abstract

Given a set of positive-weighted points and a query rectangle r (specified by a client) of given extents, the goal of a maximizing range sum (MaxRS) query is to find the optimal location of r such that the total weights of all points covered by r is maximized. In this paper, we address the problem of processing MaxRS queries over road network databases and propose two new external memory methods. Through a set of simulations, we evaluate the performance of the proposed methods.

View full abstract

The numerical error of a sample Mahalanobis distance (T²=y'S^-1y) with sample covariance matrix S is investigated. It is found that in order to suppress the numerical error of T², the following conditions need to be satisfied. First, the reciprocal square root of the condition number of S should be larger than the relative error of calculating floating-point real-number variables. The second proposed condition is based on the relative error of the observed sample vector y in T². If the relative error of y is larger than the relative error of the real-number variables, the former governs the numerical error of T². Numerical experiments are conducted to show that the numerical error of T² can be suppressed if the two above-mentioned conditions are satisfied.

View full abstract

The cycling wheelchair “Profhand” was developed in Japan as locomotion and lower limb rehabilitation device for hemiplegic subjects and elderly persons. Functional electrical stimulation (FES) control of paralyzed lower limbs enables application of the Profhand to paraplegic subjects as a rehabilitation device. In this paper, simplified muscle stimulation control for FES cycling with Profhand was examined for practical application, because cycling speed was low and not stable in our preliminary study and there was a difficulty in setting stimulation electrodes for the gluteus maximus. First, a guideline of target cycling speed to be achieved by FES cycling was determined from voluntary cycling with healthy subjects in order to evaluate FES cycling control. The cycling speed of 0.6m/s was determined as acceptable value and 1.0m/s was as ideal one. Then, stimulation to the gluteus maximus and that to the dorsiflexor muscles in addition to the quadriceps femoris were examined for simple FES cycling control for Profhand with healthy subjects. Stimulation timing was adjusted automatically during cycling based on muscle response time to electrical stimulation and cycling speed, which was shown to be effective for FES cycling control. Simple FES cycling control with Profhand removing stimulation to the gluteus maximus was found to be feasible. Stimulation to the dorsiflexor muscles with the quadriceps femoris was suggested to be effective for practical, simple FES cycling with Profhand in case of removing the gluteus maximus stimulation.

View full abstract

We propose a kernel-based quadratic classification method based on kernel principal component analysis (KPCA). Subspace methods have been widely used for multiclass classification problems, and they have been extended by the kernel trick. However, there are large computational complexities for the subspace methods that use the kernel trick because the problems are defined in the space spanned by all of the training samples. To reduce the computational complexity of the subspace methods for multiclass classification problems, we extend Oja's averaged learning subspace method and apply a subset approximation of KPCA. We also propose an efficient method for selecting the basis vectors for this. Due to these extensions, for many problems, our classification method exhibits a higher classification accuracy with fewer basis vectors than does the support vector machine (SVM) or conventional subspace methods.

View full abstract

Recovery of low-rank matrices has seen significant activity in many areas of science and engineering, motivated by theoretical results for exact reconstruction guarantees and interesting practical applications. Recently, numerous methods incorporated the nuclear norm to pursue the convexity of the optimization. However, this greatly restricts its capability and flexibility in dealing with many practical problems, where the singular values have clear physical meanings. This paper studies a generalized non-convex low-rank approximation, where the singular values are in l_p-heuristic. Then specific results are derived for image restoration, including denoising and deblurring. Extensive experimental results on natural images demonstrate the improvement of the proposed method over the recent image restoration methods.

View full abstract

This paper introduces a simple but effective way to boost the performance of scene classification through a novel approach to the LLC coding process. In our proposed method, a local descriptor is encoded not only with k-nearest visual words but also with k-farthest visual words to produce more discriminative code. Since the proposed method is a simple modification of the image classification model, it can be easily integrated into various existing BoF models proposed in various areas, such as coding, pooling, to boost their scene classification performance. The results of experiments conducted with three scene datasets: 15-Scenes, MIT-Indoor67, and Sun367 show that adding k-farthest visual words better enhances scene classification performance than increasing the number of k-nearest visual words.

View full abstract

In recent years, the need to build solid state drive (SSD)-based cloud storage systems has been increasing in order to process the big data generated by lots of Internet of Things devices and Internet users. Because these kinds of cloud systems require high performance and reliable storage, the use of flash-based Redundant Array of Independent Disks (RAID) will increase. But in flash-based RAID storage, parity data must be updated with every data write operation, which can more quickly overwhelm SSD's lifespan. To solve this problem, this letter proposes parity data deduplication for OpenStack cloud storage systems using an all flash array. Unlike the traditional data deduplication method, it only removes parity data, which will be stored in the parity disks of the all flash array. Experiments show that the proposed parity data deduplication method can efficiently reduce the number of parity data write operations, compared to the traditional data deduplication method.

View full abstract

Content centric network (CCN) is conceived as a good candidate for a futuristic Internet paradigm due to its simple and robust communication mechanism. By directly applying the CCN paradigm in wireless multihop mobile ad hoc networks, we experience various kind of issues such as packet flooding, data redundancy, packet collisions, and retransmissions etc., due to the broadcast nature of the wireless channel. To cope with the problems, in this study, we propose a novel location-aware forwarding and caching scheme for CCN-based mobile ad hoc networks. Extensive simulations are performed by using simulator, named ndnSIM. Experiment results show that proposed scheme does better as compared to other schemes in terms of content retrieval time and the number of Interest retransmissions triggered in the network.

View full abstract

The so-called numerical alphabet has been established as one of the various memorization systems. It enables numbers to be transformed into words. In that way memorizing numbers is highly alleviated, since words are to be memorized instead of numbers, which is substantially easier. In order to master the technique of transforming numbers into words (for memorizing them), as well as transforming words back to numbers, a person has to practice. Upon adopting the numerical alphabet, one then has to practice various examples and translate numbers into proper words and words into proper numbers. This paper describes the computer application we have developed that helps in this process. To our knowledge, this is the first complete application of this type ever created. We also show the results of the students' number-memorization tests, performed before and after using the application, which show significant improvements.

View full abstract

We propose a multi-label feature selection method that considers feature dependencies. The proposed method circumvents the prohibitive computations by using a low-rank approximation method. The empirical results acquired by applying the proposed method to several multi-label datasets demonstrate that its performance is comparable to those of recent multi-label feature selection methods and that it reduces the computation time.

View full abstract

In this letter, we propose a novel texture descriptor that takes advantage of an anisotropic neighborhood. A brand new encoding scheme called Reflection and Rotation Invariant Uniform Patterns (rriu2) is proposed to explore local structures of textures. The proposed descriptor is called Oriented Local Binary Patterns (OLBP). OLBP may be incorporated into other varieties of Local Binary Patterns (LBP) to obtain more powerful texture descriptors. Experimental results on CUReT and Outex databases show that OLBP not only significantly outperforms LBP, but also demonstrates great robustness to rotation and illuminant changes.

View full abstract

Vector data differs in the rasterized height field by data type. It is difficult to render dynamic vectors on height field because their shapes and locations may change at any time. This letter proposes a novel method: View-dependent Projective Atlases (VdPAs). As an intermediate data source, VdPAs act as rendering targets which enable height field and vectors to be rasterized at the same resolution. Then, VdPAs can be viewed as super-tiles. State of art height field rendering algorithms can be used for scenario rendering. Experimental results demonstrate that atlases are able to make dynamic vectors to be rendered on height field with real-time performance and high quality.

View full abstract