Performance In Real Applications Research Articles

Look-Angle-Shaping Guidance Law for Impact Angle and Time Control With Field-of-View Constraint

In this paper, we propose a guidance law that considers the field-of-view (FOV) limitation as well as the impact angle and time constraints. In order to develop the guidance law, we first design a desired look angle that satisfies the FOV limitation and impact angle constraint. Since the desired look angle is designed as a function of a gain that determines the length of the remaining trajectory, the impact time can also be controlled by adjusting the gain. The guidance law is derived so as to stabilize the actual look angle to the desired look angle based on the sliding mode control method. As a result, the proposed guidance law can intercept the stationary target at the desired impact angle and time with satisfying the FOV constraint. The proposed law is expected to achieve more accurate performance in real applications owing to its closed-loop structure with analytic formulation than the existing approach based on open-loop structure using numerical routine. The effectiveness of the proposed guidance law is demonstrated by numerical simulations.

Weight Loss for Point Clouds Classification

With the widely applying of 3D sensors such as LiDAR and RGBD camera in robots and driverless technology, 3D point clouds classification has also achieved some development. PointNet[1] is a competitive method in point clouds classification due to it’s fast speed and well performance in real applications. However, when dealing with easy-classified samples overwhelming in 3D object classification, the PointNet shows incompatibility due to its gradients are mainly determined by these easy-classified samples. In order to optimize the gradients in PointNet, we designed the SFL(Sigmoid Focal Loss) function for 3D object classification to instead of the standard softmax cross-entropy loss function. The proposed SFL function will automated refined the weights of the hard-classified samples in training, so that the new neural network can focus on these hard-classified samples to avoid the unbalance of the classifier. Our results show that the new loss function can achieve the best result in classification tasks that directly processes point clouds.

Chemical Stability of Mesoporous Oxide Thin Film Electrodes under Electrochemical Cycling: from Dissolution to Stabilization.

Mesoporous oxide thin films (MOTF) present very high surface areas and highly controlled monodisperse pores in the nanometer range. These features spurred their possible applications in separation membranes and permselective electrodes. However, their performance in real applications is limited by their reactivity. Here, we perform a basic study of the stability of MOTF toward dissolution in aqueous media using a variety of characterization techniques. In particular, we focus in their stability behavior under the influence of ionic strength, adsorption of electrochemical probes, and applied electrode potential. Mesoporous silica thin films present a limited chemical stability after electrochemical cycling, particularly under high ionic strength, due to their high specific surface area and the interactions between the electrochemical probes and the surface. In contrast, TiO2 or Si0.9Zr0.1O2 matrices present higher stability; thus, they are an adequate alternative to produce accessible, sensitive, and robust permselective electrodes or membranes that perform under a wide variety of conditions.

Semi-Flexible Material: A Solution for High-Performance Pavement Infrastructures

Asphalt pavement technology has been widely adopted due its good mechanical performance, its excellent surface properties, the ease of construction and driving comfort. Nevertheless, asphalt pavements, commonly also known as “flexible pavements” have demonstrated to present high moisture susceptibility, low resistance to several chemical agents and poor long-term performance. Moreover, both high and low temperatures can produce different damages in flexible pavement such as rutting and cracking. On the other hand, rigid pavements have been losing popularity mainly due to the presence of transversal joints, required to consider the expansion of concrete slabs that produce a reduction of user-comfort and safety. In addition, Concrete pavement need longer construction procedure and a curing time before opening to traffic. Hence, a new technology has been developed in the last years in order to gather the essential properties that characterize rigid and flexible pavements. This technology, usually called semi-flexible pavements, or grouted macadam, consist in producing a highly open porous asphalt mix (voids around 25-30%) and filling the voids with selected cementitious grout. Thus, the ease and speed of construction of flexible pavement, joint with the optimal surface properties is combined with the concrete good mechanical behaviour, resistance to chemical agents and limited temperature susceptibility. The final product is a high performance pavement material, particularly indicated to high traffic volume roads, airports, ports, and industrial areas. However, the fact that this type of pavement is still innovative compared to rigid and flexible pavement has made that there is no standard procedure to produce this type of mixture. For this reason, the characteristics of the mixtures vary with authors and in the different real scale applications. In this manner, this paper aims at studying the most significant advances in the area of semi-flexible pavements, describing, comparing and statistically analysing the different investigation and practical experiences. Mix-design details, i.e., asphalt binder characteristics, aggregate properties and cementitious grout composition are deeply investigated, as well as mixing and compaction procedures, lab testing and real application performance of this innovative and promising pavement material.

Hardware/Software Adaptive Cryptographic Acceleration for Big Data Processing

Along with the explosive growth of network data, security is becoming increasingly important for web transactions. The SSL/TLS protocol has been widely adopted as one of the effective solutions for sensitive access. Although OpenSSL could provide a freely available implementation of the SSL/TLS protocol, the crypto functions, such as symmetric key ciphers, are extremely compute-intensive operations. These expensive computations through software implementations may not be able to compete with the increasing need for speed and secure connection. Although there are lots of excellent works with the objective of SSL/TLS hardware acceleration, they focus on the dedicated hardware design of accelerators. Hardly of them presented how to utilize them efficiently. Actually, for some application scenarios, the performance improvement may not be comparable with AES-NI, due to the induced invocation cost for hardware engines. Therefore, we proposed the research to take full advantages of both accelerators and CPUs for security HTTP accesses in big data. We not only proposed optimal strategies such as data aggregation to advance the contribution with hardware crypto engines, but also presented an Adaptive Crypto System based on Accelerators (ACSA) with software and hardware codesign. ACSA is able to adopt crypto mode adaptively and dynamically according to the request character and system load. Through the establishment of 40 Gbps networking on TAISHAN Web Server, we evaluated the system performance in real applications with a high workload. For the encryption algorithm 3DES, which is not supported in AES-NI, we could get about 12 times acceleration with accelerators. For typical encryption AES supported by instruction acceleration, we could get 52.39% bandwidth improvement compared with only hardware encryption and 20.07% improvement compared with AES-NI. Furthermore, the user could adjust the trade-off between CPU occupation and encryption performance through MM strategy, to free CPUs according to the working requirements.

Asymmetric Deep Supervised Hashing

Hashing has been widely used for large-scale approximate nearest neighbor search because of its storage and search efficiency. Recent work has found that deep supervised hashing can significantly outperform non-deep supervised hashing in many applications. However, most existing deep supervised hashing methods adopt a symmetric strategy to learn one deep hash function for both query points and database (retrieval) points. The training of these symmetric deep supervised hashing methods is typically time-consuming, which makes them hard to effectively utilize the supervised information for cases with large-scale database. In this paper, we propose a novel deep supervised hashing method, called asymmetric deep supervised hashing (ADSH), for large-scale nearest neighbor search. ADSH treats the query points and database points in an asymmetric way. More specifically, ADSH learns a deep hash function only for query points, while the hash codes for database points are directly learned. The training of ADSH is much more efficient than that of traditional symmetric deep supervised hashing methods. Experiments show that ADSH can achieve state-of-the-art performance in real applications.

Performance exploration of temperature swing adsorption technology for carbon dioxide capture

Adsorption technology is recognised to be a promising CO2 capture method due to its desirable characteristics e.g. reusable nature of adsorbents, low capital investment and easy automatic operation. To further improve thermal performance, internal heat recovery is adopted for adsorption CO2 capture through analogy with adsorption refrigeration. Based on carbon pump theory, thermal performance of 4-step temperature swing adsorption (TSA) processes is analysed at various adsorption/desorption temperatures and pressures. Exergy efficiency of adsorption CO2 capture with and without heat recovery will be evaluated and compared by using experimental adsorption characteristics of activated carbon. Metal part and unused percentage of adsorption reactor are defined to further assess their influence on system performance in real application. Results indicate that sensible heat of adsorbents and adsorbed phase account for the major part of heat consumption. For different desorption/adsorption temperatures and pressures, theoretical exergy efficiency of 4-step TSA cycle ranges from 0.022 to 0.221. Heat recovery is conducive to exergy efficiency. Through heat recovery, exergy efficiency could be improved from 54.3% to 84.6% when mass ratio increases from 0 to 8. Similarly, the improvement by using heat recovery is up to 90% in terms of different unused percentages.

Runtime Techniques to Mitigate Soft Errors in Network-on-Chip (NoC) Architectures

As aggressive scaling continues to push multiprocessor system-on-chips (MPSoCs) to new limits, complex hardware structures combined with stringent area and power constraints will continue to diminish reliability. Waning reliability in integrated circuits will increase the susceptibility of transient and permanent faults. There is an urgent demand for adaptive error correction coding (ECC) schemes in network-on-chips to provide fault tolerance and improve overall resiliency of MPSoC architectures. The goal of adaptive ECC schemes should be to maximize power savings when faults are infrequent and increase application speedup by boosting fault coverage when faults are frequent. In this paper, we propose runtime adaptive scrubbing (RAS), a novel multilayered error correction and detection scheme with three modes of operation enabled by an area-efficient configurable encoder for encoding packets on the switch-to-switch (s2s) layer, thus preventing faults from accumulating up the network stack and onto the end-to-end layer. As fault rates fluctuate we propose a dynamic methodology for improving fault localization and intelligently adapt fault coverage on demand to sustain graceful network degradation. RAS successfully improves network resiliency, fault localization, and fault coverage as compared to traditional static s2s schemes. Simulation results demonstrate that static RAS improves network speedup by 10% for Splash-2/PARSEC benchmarks on a $8 \times 8$ mesh network while reducing area overhead by 14% and incurring on an average 6.6% power penalty by boosting fault tolerance when fault rates increase. Further, our dynamic RAS scheme maintains 97.88% of network performance for real applications while incurring 20% power penalty.

Exfoliation of Few-Layer Black Phosphorus in Low-Boiling-Point Solvents and Its Application in Li-Ion Batteries

The liquid-phase exfoliation (LPE) of black phosphorus (BP) is a strategic route for the large-scale production of phosphorene and few-layer BP (FL-BP) flakes. The exploitation of this exfoliated material in cutting-edge technologies, e.g., in flexible electronics and energy storage, is however limited by the fact that the LPE of BP is usually carried out at a high boiling point and in toxic solvents. In fact, the solvent residual is detrimental to device performance in real applications; thus, complete solvent removal is critical. Here, we tackle these issues by exfoliating BP in different low boiling-point solvents. Among these solvents, we find that acetone also provides a high concentration of exfoliated BP, leading to the production of FL-BP flakes with an average lateral size and thickness of c.a. 30 and 7 nm, respectively. The use of acetone to produce less defective few-layer BP flakes (FL-BPacetone) from bulk crystals is a straightforward process which enables the rapid preparation of homogeneous thin films thanks to the fast solvent evaporation. The ratio of edge to bulk atoms for the BP flakes here produced, combined with the use of low-boiling-point solvents for the exfoliation process suggests that these thin films are promising anodes for lithium-ion batteries. To this end, we tested Li-ion half cells with FL-BPacetone anodes achieving a reversible specific capacity of 480 mAh/g at a current density of 100 mA/g, over 100 charge/discharge cycles. Moreover, a reversible specific capacity of 345 mAh/g is achieved for the FL-BPacetone-based anode at a high current density (i.e., 1 A/g). These findings indicate that the FL-BPacetone-based battery is promising with regards to the design of fast charge/discharge devices. Overall, the presented process is a step forward toward the fabrication of phosphorene-based devices.

Implementation of viscoelastic mud-induced energy attenuation in the third-generation wave model, SWAN

The interaction of waves with fluid mud can dissipate the wave energy significantly over few wavelengths. In this study, the third-generation wave model, SWAN, was advanced to include attenuation of wave energy due to interaction with a viscoelastic fluid mud layer. The performances of implemented viscoelastic models were verified against an analytical solution and viscous formulations for simple one-dimensional propagation cases. Stationary and non-stationary test cases in the Surinam coast and the Atchafalaya Shelf showed that the inclusion of the mud-wave interaction term in the third-generation wave model enhances the model performance in real applications. A high value of mud viscosity (of the order of 0.1 m2/s) was required in both field cases to remedy model overestimation at high frequency ranges of the wave spectrum. The use of frequency-dependent mud viscosity value improved the performance of model, especially in the frequency range of 0.2–0.35 Hz in the wave spectrum. In addition, the mud-wave interaction might affect the high frequency part of the spectrum, and this part of the wave spectrum is also affected by energy transfer from wind to waves, even for the fetch lengths of the order of 10 km. It is shown that exclusion of the wind input term in such cases might result in different values for parameters of mud layer when inverse modeling procedure was employed. Unlike viscous models for wave-mud interaction, the inverse modeling results to a set of mud parameters with the same performance when the viscoelastic model is used. It provides an opportunity to select realistic mud parameters which are in more agreement with in situ measurements.

TiO2/RbPbI3 halide perovskite solar cells

Inorganic-organic halide perovskites hold the great promise for next-generation photovoltaics due to their excellent high performance and low cost. However, major limitation for the commercialization of perovskite solar cell can be attributed to the transformation and degradation of lead halide perovskite during the exposure of environmental humidity and photon irradiation. To solve these problems, herein, we apply the one-dimensional and inorganic RbPbI3 perovskite into the solar cell because it shows the superior stability in environmental conditions. After RbPbI3 perovskite was applied into the solar cell with a FTO/TiO2/RbPbI3/Spiro-MeOTAD/Au configuration, the device exhibits an open circuit voltage of 0.62V, photocurrent density of 3.75mA/cm2, fill factor of 44.60%, and 1.04% of PCE by reverse sweeping direction. Even though the performance of RbPbI3 device was still lower than other perovskite solar cells, this approach enabled us to establish the key step to make a highly stable perovskite film, leading to the best photovoltaic performance for real applications.

Design new hole transport materials for efficient perovskite solar cells by suitable combination of donor and core groups

The electronic, optical, and hole transporting mobility of three organic hole transporting materials (HTMs), X59, X59-P, and X59-T, are investigated by combination of first principle and molecular dynamics associated with Marcus theory and Einstein equation. As compared with the experimental reported X59, the new designed X59-P has more stable HOMO energy level. Moreover, the latter has smaller reorganization energy and larger hole transfer integral resulting in the larger hole transporting mobility. Besides the hole transporting mobility, the solubility and stability of two designed molecules are also evaluated by comparison with X59, which are two important items to determine the cost and performance in real application of solar cell. More importantly, they would be synthesized in a benign condition without expensive materials. Our studies introduce a possible pathway to explore the efficient HTMs by suitable combination mode rather than development of new groups.

Study of mass transfer coefficient in membrane desalination

Membrane desalination (MD) is a thermally-driven separation process. The driving force in the MD process is the vapor pressure difference induced by the temperature difference across a hydrophobic membrane as a barrier. MD is a potential route to obtain high-quality fresh water from sea water or brine by using low-grade energy. It is important to estimate and predict the MD performance especially in terms of permeate flux under given conditions. This is very meaningful in MD plications. In this work, the MD performance was studied not only by varying the feed temperature but also by varying the coolant temperature in a lab scale AGMD system. The permeate flux and salt reject ratio were tested under various conditions. The mass transfer coefficient in porous membranes was studied in detail both theoretically and experimentally. A new approach to estimate the permeate flux was proposed based on the study of the mass transfer coefficient in MD systems. The estimation of MD performance in real applications can be much simplified based on this study.

Decentralized Clustering by Finding Loose and Distributed Density Cores

Centroid-based clustering approaches fail to recognize extremely complex patterns that are non-isotropic. We analyze the underlying causes and find some inherent flaws in these approaches, including Shape Loss, False Distances and False Peaks, which typically cause centroid-based approaches to fail when applied to complex patterns. As an alternative to current methods, we propose a hybrid decentralized approach named DCore, which is based on finding density cores instead of centroids, to overcome these flaws. The underlying idea is that we consider each cluster to have a shrunken density core that roughly retains the shape of the cluster. Each such core consists of a set of loosely connected local density peaks of higher density than their surroundings. Borders, edges and outliers are distributed around the outsides of these cores in a hierarchical structure. Experiments demonstrate that the promise of DCore lies in its power to recognize extremely complex patterns and its high performance in real applications, for example, image segmentation and face clustering, regardless of the dimensionality of the space in which the data are embedded.

Surface specific asperity model for prediction of friction in boundary and mixed regimes of lubrication

Machine downsizing, increased loading and better sealing performance have progressively led to thinner lubricant films and an increased chance of direct surface interaction. Consequently, mixed and boundary regimes of lubrication are prevalent with ubiquitous asperity interactions, leading to increased parasitic losses and poor energy inefficiency. Surface topography has become an important consideration as it influences the prevailing regime of lubrication. As a result a plethora of machining processes and surface finishing techniques have emerged. The stochastic nature of the resulting topography determines the separation at which asperity interactions are initiated and ultimately affect the conjunctional load carrying capacity and operational efficiency. The paper presents a procedure for modelling of asperity interactions of real rough surfaces, from measured data, which do not conform to the usually assumed Gaussian distributions. The model is validated experimentally using a bench top reciprocating sliding test rig. The method demonstrates accurate determination of the onset of mixed regime of lubrication. In this manner, realistic predictions are made for load carrying and frictional performance in real applications where commonly used Gaussian distributions can lead to anomalous predictions.

Fast Asynchronous Parallel Stochastic Gradient Descent: A Lock-Free Approach with Convergence Guarantee

Stochastic gradient descent (SGD) and its variants have become more and more popular in machine learning due to their efficiency and effectiveness. To handle large-scale problems, researchers have recently proposed several parallel SGD methods for multicore systems. However, existing parallel SGD methods cannot achieve satisfactory performance in real applications. In this paper, we propose a fast asynchronous parallel SGD method, called AsySVRG, by designing an asynchronous strategy to parallelize the recently proposed SGD variant called stochastic variance reduced gradient (SVRG). AsySVRG adopts a lock-free strategy which is more efficient than other strategies with locks. Furthermore, we theoretically prove that AsySVRG is convergent with a linear convergence rate. Both theoretical and empirical results show that AsySVRG can outperform existing state-of-the-art parallel SGD methods like Hogwild! in terms of convergence rate and computation cost.

T-MQM: Testbed based Multi-metric Quality Measurement of Sensor Deployment for Precision Agriculture-A Case Study

Efficient sensor deployment is one of the primary requirements of the precision agriculture use case of wireless sensor networks (WSNs) to provide qualitative and optimal coverage and connectivity. The application-based performance variations of the geometrical-model-based sensor deployment patterns restrict the generalization of a specific deployment pattern for all applications. Furthermore, single or double metrics-based evaluation of the deployment patterns focusing on theoretical or simulation aspects can be attributed to the difference in performance of real applications and the reported performance in the literature. In this context, this paper proposes a testbed-based multi-metric quality measurement of sensor deployment for the precision agriculture use case of WSNs. Specifically, seven metrics are derived for the qualitative measurement of sensor deployment patterns for precision agriculture. The seven metrics are quantified for four sensor deployment patterns to measure the quality of coverage and connectivity. Analytical- and simulation-based evaluations of the measurements are validated through testbed experiment-based evaluations which are carried out in “INDRIYA” WSNs testbed. Toward realistic research impact, the investigative evaluation of the geometrical-model-based deployment patterns presented in this paper could be useful for practitioners and researchers in developing performance guaranteed applications for precision agriculture and novel coverage and connectivity models for deployment patterns.

IP packet interleaving for UDP bursty losses

The bursty nature of losses over the Internet is constantly asking for effective solutions. In this work, we use a comprehensive approach to study packet interleaving for coping with loss burstiness. Our aim is to assess if and how packet interleaving can be employed in real networks at the IP layer and what benefits real UDP-based applications may expect. Through an analytical study we determine the interleaving configuration to be used as a function of the loss characteristics. Then, in simulation we confirm these findings and highlight also potential counter effects. Afterwards, we design and develop a real application for packet interleaving. We move a step ahead towards real networks using a testbed that comprises an emulated WAN. Thanks to it, we study and solve a number of issues arising in real environments such as network dynamics and interleaving performance. Finally, we deploy the packet interleaver in the wild and we study the improvements achievable in general and by a real application. Our work shows that providing benefits to real applications is not an easy task. However, our packet interleaver can effectively decorrelate losses at the IP layer in real networks and highly increase real application performance, for example, video distortion can be reduced 65%.

A unified approach for detecting and eliminating selfish nodes in MANETs using TBUT

Abstract Recent years have witnessed the increasing efforts toward making architecture standardization for the secured wireless mobile ad hoc networks. In this scenario when a node actively utilizes the other node resources for communicating and refuses to help other nodes in their transmission or reception of data, it is called a selfish node. As the entire mobile ad hoc network (MANETs) depends on cooperation from neighboring nodes, it is very important to detect and eliminate selfish nodes from being part of the network. In this paper, token-based umpiring technique (TBUT) is proposed, where every node needs a token to participate in the network and the neighboring nodes act as umpire. This proposed TBUT is found to be very efficient with a reduced detection time and less overhead. The security analysis and experimental results have shown that TBUT is feasible for enhancing the security and network performance of real applications.

A survey on fingerprint minutiae-based local matching for verification and identification: Taxonomy and experimental evaluation

Fingerprint recognition has found a reliable application for verification or identification of people in biometrics. Globally, fingerprints can be viewed as valuable traits due to several perceptions observed by the experts; such as the distinctiveness and the permanence on humans and the performance in real applications. Among the main stages of fingerprint recognition, the automated matching phase has received much attention from the early years up to nowadays. This paper is devoted to review and categorize the vast number of fingerprint matching methods proposed in the specialized literature. In particular, we focus on local minutiae-based matching algorithms, which provide good performance with an excellent trade-off between efficacy and efficiency. We identify the main properties and differences of existing methods. Then, we include an experimental evaluation involving the most representative local minutiae-based matching models in both verification and evaluation tasks. The results obtained will be discussed in detail, supporting the description of future directions.