Communication costs have become a performance bottleneck in many applications, and are a big issue for high performance computing on massively parallel machines. This paper proposes a halo exchange method for unstructured sparse matrix vector products within the algebraic multigrid method, and evaluate it on a supercomputer with mesh/torus networks. In our numerical tests with a Poisson problem, the proposed method accelerates the linear solver more than 14 times with 23040 cores.

In typical end-to-end recovery protocols, an ACK segment is delivered to a source node over a single path. ACK loss requires the source to retransmit the corresponding data packet. However, in underwater wireless sensor networks which prefer flooding-based routing protocols, the source node has redundant chances to receive the ACK segment since multiple copies of the ACK segment can arrive at the source node along multiple paths. Since existing RTO calculation algorithms do not consider inherent features of underlying routing protocols, spurious packet retransmissions are unavoidable. Hence, in this letter, we propose a new ACK loss-aware RTO calculation algorithm, which utilizes statistical ACK arrival times and ACK loss rate, in order to reduce such retransmissions.

Post-Silicon Tunable (PST) buffers are widely adopted in high-performance integrated circuits to fix timing violations introduced by process variations. In typical optimization procedures, the statistical timing analysis of the circuits with PST clock buffers will be executed more than 2000 times for large scale circuits. Therefore, the efficiency of the statistical timing analysis is crucial to the PST clock buffer optimization algorithms. In this paper, we propose a stochastic collocation based efficient statistical timing analysis method for circuits with PST buffers. In the proposed method, we employ the Howard algorithm to calculate the clock periods of the circuits on less than 100 deterministic sparse-grid collocation points. Afterwards, we use these obtained clock periods to derive the yield of the circuits according to the stochastic collocation theory. Compared with the state-of-the-art statistical timing analysis method for the circuits with PST clock buffers, the proposed method achieves up to 22X speedup with comparable accuracy.

A test structure for charge-based capacitance measurement (CBCM) method has been developed to evaluate the threshold voltage variability from capacitance-voltage (C-V) curves of actual size metal-oxide-semiconductor field-effect-transistors (MOSFETs). The C-V curves from accumulation to inversion are measured for the MOSFETs having various channel dimensions using this test structure. Intrinsic capacitance components between the MOSFET electrodes are extracted from those C-V curves which are considered to include parasitic capacitance component. The intrinsic C-V curves are used for attempting to extract threshold voltage variations of their MOSFETs. It is found that the developed test structure is very useful for the evaluation of MOSFETs variability, because the derivation in MOSFET C-V curves is not influenced by current measurement noise.

This paper presents a design of the software and hardware modules of an embedded board with both a sensor and an interface circuit which not only control home light-emitting diode (LED) lighting appliances but also reduce their differences in brightness caused by luminous decay. This design consists of four parts: an automatically adjusted LED driver, environment illumination detection, a wireless remote control unit and an automatic brightness control.

This paper proposes a background noise estimation method using an outer product expansion with non-linear filters for ELF (extremely low frequency) electromagnetic (EM) waves. We proposed a novel source separation technique that uses a tensor product expansion. This signal separation technique means that the background noise, which is observed in almost all input signals, can be estimated using a tensor product expansion (TPE) where the absolute error (AE) is used as the error function, which is thus known as TPE-AE. TPE-AE has two problems: the first is that the results of TPE-AE are strongly affected by Gaussian random noise, and the second is that the estimated signal varies widely because of the random search. To solve these problems, an outer product expansion based on a modified trimmed mean (MTM) is proposed in this paper. The results show that this novel technique separates the background noise from the signal more accurately than conventional methods.

This paper proposes a reliable data aggregation scheduling that uses caching and re-transmission based on track topology. In the proposed scheme, a node detects packet losses by overhearing messages that includes error indications of the child nodes, from its neighbor nodes. If packet losses are detected, as a backup parent, the node retransmits the lost packet. A retransmission strategy is added into the adaptive timeout scheduling scheme, which adaptively configures both the timeout and the collection period according to the potential level of an event occurrence. The retransmission steps cause an additional delay and power consumption of the sensor nodes, but dramatically increase the data accuracy of the aggregation results. An extensive simulation under various workloads shows that the proposed scheme outperforms other schemes in terms of data accuracy and energy consumption.

This letter proposes an efficient lossless compression method for high dynamic range (HDR) images in OpenEXR format. The proposed method transforms an HDR image to an indexed image and packs the histogram of the indexed image. Finally the packed image is losslessly compressed by using any existing lossless compression algorithm such as JPEG 2000. Experimental results show that the proposed method reduces the bit rate of compressed OpenEXR images compared with equipped lossless compression methods of OpenEXR format.

In this paper, the H2/H∞ control problem for a class of stochastic discrete-time linear systems with state-, control-, and external-disturbance-dependent noise or (x, u, v)-dependent noise involving multiple decision makers is investigated. It is shown that the conditions for the existence of a strategy are given by the solvability of cross-coupled stochastic algebraic Riccati equations (CSAREs). Some algorithms for solving these equations are discussed. Moreover, weakly-coupled large-scale stochastic systems are considered as an important application, and some illustrative examples are provided to demonstrate the effectiveness of the proposed decision strategies.

The amplitude damping (AD) quantum channel is one of the models describing evolution of quantum states. The construction of quantum error correcting codes for the AD channel based on classical codes has been presented, and Shor et al. proposed a class of classical codes over F3 which are efficiently applicable to this construction. In this study, we expand Shor's construction to that over F7, and succeeded to construct an AD code that has better parameters than AD codes constructed by Shor et al.

Network virtualization is an essential technology for cloud datacenters that provide multi-tenancy services. SDN-enabled datacenters have introduced an edge-overlay (distributed tunneling) model to construct virtual tenant networks. The edge-overlay model generally uses L2-in-L3 tunneling protocols like VXLAN. However, the tunneling-based edge-overlay model has some performance and compatibility problems. We have proposed a yet another overlay approach without using IP tunneling. Our model leverages two methods, OpenFlow-based Virtual/Physical MAC address translation and host-based VLAN ID usage. The former method replaces VMs' MAC addresses to physical servers' ones, which prevents frame encapsulation as well as unnecessary MAC address learning by physical switches. The later method breaks a limitation of the number of VLAN-based virtual tenant networks (4094) by allocating entire VLAN ID space to each physical server and by mapping VLAN ID to VM with OpenFlow controller support. In our model, any special hardware equipment like OpenFlow hardware switches is not required and only software-based virtual switches and the controller are used. In this paper, we evaluated the performance of the proposed model comparing with the tunneling model using 40GbE environment. The results show that the performance of VM-to-VM communication with the proposed model is close to that of physical communication and exceeds 10Gbps throughput with large TCP segment, and the proposed model shows better scalability for the number of VMs.

In recent years, Ethernet fabrics have been developed with a view to using resources efficiently and simplifying the operation of data center networks. With Ethernet fabrics, frames are forwarded along the shortest paths based on routing tables without blocking ports. Ethernet fabrics are expected to be employed in more general networks including carrier access networks. In particular, the use of shortest path bridging MAC (SPBM) is expected to allow smooth migration from existing networks. With SPBM, networks can be flexibly constructed on demand in any network topology. If an arbitrary topology is constructed, traffic paths can overlap on specific links and throughput unfairness occurs. However, it is difficult to achieve accurate weighted fairness with existing schemes. This paper proposes employing weighted N rate N+1 color marking (WNRN+1CM) in SPBM networks to achieve per-flow weighted fairness. WNRN+1CM was developed to realize weighted fairness in layer-2 ring networks and the applicability to other network topologies has not yet been discussed. The outline of WNRN+1CM in SPBM is as follows. The weight and the maximum rate are provided for each flow at edge bridges. When edge bridges receive frames from outside the SPBM domain, they assign colors to frames according to the input rate and the weight of each flow. The color indicates the dropping priority. If the input rate exceeds the maximum rate, frames are discarded to limit the throughput. Core bridges selectively discard frames based on their color and the dropping threshold when congestion occurs. The bandwidth is allocated based on the weights. The performance of WNRN+1CM is evaluated with a theoretical analysis and computer simulations. WNRN+1CM can achieve weighted fairness in aggregation networks and multipoint networks. The throughput ratio matches the weights and the flow throughputs are limited to their maximum rate regardless of changes in traffic.

In 3GPP (3-rd Generation Partnership Project) LTE (Long Term Evolution) systems, D2D (Device-to-Device) communication has been selected as a next generation study item. In uplink D2D communication that underlies LTE systems, uplink interference signals generated by CUE (Cellular User Equipment) have a profound impact on the throughput of DUE (D2D User Equipment). For that reason, various resource allocation algorithms which consider interference channels have been studied; however, these algorithms assume accurate channel estimation and feedback of D2D related links. Therefore, in order to estimate uplink channels of D2D communication, SRS (Sounding Reference Signal) defined in LTE uplink channel structure can be considered. However, when the number of interferes increases, the SRS based method incurs significant overheads such as side information, operational complexity, channel estimation and feedback to UE. Therefore, in this paper, we propose an efficient channel estimation and CSI (Channel State Information) feedback method for D2D communication, and its application in LTE systems. We verify that the proposed method can achieve a similar performance to SRS based method with lower operational complexity and overhead.

The MR image segmentation is always a challenging problem because of the intensity inhomogeneity. Many existing methods don't reach their expected segmentations; besides their implementations are usually complicated. Therefore, we originally interleave the extended Otsu segmentation with bias field estimation in an energy minimization. Via our proposed method, the optimal segmentation and bias field estimation are achieved simultaneously throughout the reciprocal iteration. The results of our method not only satisfy the required classification via its applications in the synthetic and the real images, but also demonstrate that our method is superior to the baseline methods in accordance with the performance analysis of JS metrics.

In this paper, we investigate the problems of the established congestion solution and then introduce a self-adjustable rate control that supports quality of service assurances over multi-hop wireless mesh networks. This scheme eliminates two phases of the established congestion solution and works on the MAC layer for congestion control. Each node performs rate control by itself so network congestion is eliminated after it independently collects its vector parameters and network status parameters for rate control. It decides its transmission rate based on a predication model which uses a rate function including a congestion risk level and a passing function. We prove that our scheme works efficiently without any negative effects between the network layer and the data link layer. Simulation results show that the proposed scheme is more effective and has better performance than the existing method.

This paper proposes a novel method for determining a three-dimensional (3D) bounding box to estimate pose (position and orientation) and size of a 3D object corresponding to a segmented object region in an image acquired by a single calibrated camera. The method is designed to work upon an object on the ground and to determine a bounding box aligned to the direction of the object, thereby reducing the number of degrees of freedom in localizing the bounding box to 5 from 9. Observations associated with the structural properties of back-projected object regions on the ground are suggested, which are useful for determining the object points expected to be on the ground. A suitable base is then estimated from the expected on-ground object points by applying to them an assumption of bilateral symmetry. A bounding box with this base is finally constructed by determining its height, such that back-projection of the constructed box onto the ground minimally encloses back-projection of the given object region. Through experiments with some 3D-modelled objects and real objects, we found that a bounding box aligned to the dominant direction estimated from edges with common direction looks natural, and the accuracy of the pose and size is enough for localizing actual on-ground objects in an industrial working space. The proposed method is expected to be used effectively in the fields of smart surveillance and autonomous navigation.

In this paper, an adaptive expansion strategy (AES) is proposed for multiple-input/multiple-output (MIMO) detection in the presence of circular signals. By exploiting channel properties, the AES classifies MIMO channels into three types: excellent, average and deep fading. To avoid unnecessary branch-searching, the AES adopts single expansion (SE), partial expansion (PE) and full expansion (FE) for excellent channels, average channels and deep fading channels, respectively. In the PE, the non-circularity of signal is exploited, and the widely linear processing is extended from non-circular signals to circular signals by I (or Q) component cancellation. An analytical performance analysis is given to quantify the performance improvement. Simulation results show that the proposed algorithm can achieve quasi-optimal performance with much less complexity (hundreds of flops/symbol are saved) compared with the fixed-complexity sphere decoder (FSD) and the sphere decoder (SD).

In this paper, we present an efficient time-of-arrival (TOA)-based localization method for wireless sensor networks. The goal of a localization system is to accurately estimate the geographic location of a wireless device. In real wireless sensor networks, accurately estimating mobile device location is difficult because of the presence of various errors. Therefore, localization methods have been studied in recent years. In indoor environments, the accuracy of wireless localization systems is affected by non-line-of-sight (NLOS) errors. The presence of NLOS errors degrades the performance of wireless localization systems. In order to effectively estimate the location of the mobile device, NLOS errors should be recognized and mitigated in indoor environments. In the TOA-based ranging method, the distance between the two wireless devices can be computed by multiplying a signal's propagation delay time by the speed of light. TOA-based localization measures the distance between the mobile station (MS) and three or more base stations (BSs). However, each of the NLOS errors of the measured distance between the i-th BS and the MS is different due to dissimilar obstacles in the direct signal path between the two nodes. In order to accurately estimate the location in a TOA-based localization system, an optimized localization algorithm that selects three measured distances with fewer NLOS errors is necessary. We present an efficient TOA-based localization scheme that combines three selected BSs in wireless sensor networks. This localization scheme yields improved localization performance in wireless sensor networks. In this paper, performance tests are performed, and the simulation results are verified through comparisons between various localization methods and the proposed method. As a result, proposed localization scheme using BS selection achieves remarkably better localization performance than the conventional methods. This is verified by experiments in real environments, and demonstrates a performance analysis in NLOS environments. By using BS selection, we will show an efficient and effective TOA-based localization scheme in wireless sensor networks.

There exist two intrinsic issues in multiple-shot person re-identification: (1) large differences in camera view, illumination, and non-rigid deformation of posture that make the intra-class variance even larger than the inter-class variance; (2) only a few training data that are available for learning tasks in a realistic re-identification scenario. In our previous work, we proposed a local distance comparison framework to deal with the first issue. In this paper, to deal with the second issue (i.e., to derive a reliable distance metric from limited training data), we propose an adaptive learning method to learn an adaptive distance metric, which integrates prior knowledge learned from a large existing auxiliary dataset and task-specific information extracted from a much smaller training dataset. Experimental results on several public benchmark datasets show that combined with the local distance comparison framework, our adaptive learning method is superior to conventional approaches.

Re-identifying the same person in different images is a distinct challenge for visual surveillance systems. Building an accurate correspondence between highly variable images requires a suitable dissimilarity measure. To date, most existing measures have used adapted distance based on a learned metric. Unfortunately, real-world human image data, which tends to show large intra-class variations and small inter-class differences, continues to prevent these measures from achieving satisfactory re-identification performance. Recognizing neighboring distribution can provide additional useful information to help tackle the deviation of the to-be-measured samples, we propose a novel dissimilarity measure from the neighborhood-wise relative information perspective, which can deliver the effectiveness of those well-distributed samples to the badly-distributed samples to make intra-class dissimilarities smaller than inter-class dissimilarities, in a learned discriminative space. The effectiveness of this method is demonstrated by explanation and experimentation.