Abstract: Large-scale parallel simulation applications, characterized by complex computation, vast amount of data, great storage, large quantities of samples and strong causality, can only be realized by employing highperformance simulation supercomputer. This paper proposes the idea of “blending the general and the specialized” in developing high-performance simulation supercomputer, in which various functions are enhanced, such as the compute node, specialized simulation synchronization network, globally shared I/O, and visualization system, based on the general supercomputer. The disadvantages of the HLA standard are also analyzed and the necessity is explained to develop the efficient parallel simulation infrastructure according to the characteristics of the highperformance simulation supercomputer. This paper also presents the experimental results of large-scale simulation applications running in our high-performance simulation supercomputer. As shown by such results, the execution efficiency is greatly improved in large-scale parallel simulation applications with enhanced high-performance simulation supercomputer and customized parallel simulation infrastructure.
Abstract: In this paper, we try to accelerate sparse LU factorization on GPU. We present a tiled storage format and a parallel algorithm to improve the memory access pattern, and a register blocking method to compress the on-chip working set. The OPENMP implementation of our algorithm gives more stable performance over different matrices, and outperforms SuperLU and KLU by 1.88～6 times on an Intel 8-core CPU (Central processing unit) for matrices from the Florida matrix collection. Based on this algorithm, we further propose a GPU-CPU hybrid pipelined scheme to overlap computations on CPU with computations on GPU. Compared to the better of SuperLU and KLU on an Intel 8-core CPU, our algorithm achieves 1.1～19.7-fold speedup on GPU for double precision. Compared to the OPENMP implementation of our algorithm on an Intel 8-core CPU, our GPU implementation gets a 2-fold speedup for the best cases.
Abstract: This paper presents a method which applies software trustworthiness measures to the dynamic behavior feature datasets generated at software running time compared with the static attribute feature datasets generated at software testing time in order to make recommendations for users in services selection time under the environment of SaaS. The measurement method is carried out in three stages: firstly, defining the concept of trust, software trustworthiness, static and dynamic feature datasets with fundamental calculating criteria; secondly, providing a group of formulas to illustrate congruence measurement approach for comparing the two types of feature datasets; lastly, giving the trustworthiness concept hierarchies to evaluate software trustworthiness.
Abstract: A 24GHz Voltage-controlled oscillator (VCO) was designed and fabricated in 90nm CMOS technology. New features of this VCO include a tail current source to reduce the power consumption and a noise filter to prevent phase noise degradation. Measurement of the finished chip shows that it has a tuning range of 22.5– 26.3GHz (16.9%) and achieves a phase noise performance of ?103dBc/Hz at 1MHz offset. The entire circuit of VCO with output buffer consumes 8mA current from 1.2V power supply. The FOM of the VCO is ?180dB.
Abstract: Fuzz testing has some obvious disadvantages: (1) It is ineffective with if-else statement; (2) it can hardly deal with the input with unknown fixed structure; (3) randomly generated input can hardly pass the verification of input integrity applied by the target systems. On the other hand, Symbolic execution can overcome some of the obstacles above. However, most of them didn't handle the function of verification about the input integrity of program. Due to the limitation of fuzz test and symbolic execution to the verification of input integrity, we propose a reversed taint tracing approach to overcome this problem. The key idea is that it traverses the entire path of program by symbolic execution and generates results from the path-reversed files. We found 7 unhandled exceptions in 7-zip, and revised 16 malformed files with correct CRCs to pass the 7-zip test and decompression.
Abstract: When nodes of Distributed file system are extended over wide area network, network communication has a great influence on the node selection of Distributed file system. In this paper, an improved algorithm is proposed to decrease the transport time by reducing the scale of nodes. This algorithm adopts the law of universal gravitation, which gives strategy of node movement. Meanwhile, to overcome premature or local-best solution, the theory of overcoming premature is referred, and then node can depart for a more suitable cluster. Theoretical proof shows the algorithm converges and has the top limit in the time complexity. Furthermore, experiment results give the availability and efficiency of the algorithm.
Abstract: As secure infrastructural software, secure operating systems are well known for its protection against kinds of threats and attacks. In this paper, we report the work of building EARTH operating system towards to “structurized protection” level of national security standards. EARTH has a flexible architecture supporting dynamic multiple policies, effective mandatory access control mechanism. We give formal system specification and proof of the security models and provide covert channel analysis and mitigation methods. Our experiments show that EARTH has good system performance. Furthermore, the research and development experiences and lessons learned in EARTH project is discussed.
Abstract: For the purpose of document interoperability testing, features used to measure the interoperability are defined and are associated with the concepts in ontology. A measurement model was established where document interoperability are measured according to the similarity of these features which in turn are determined by the similarity of concepts. The model can be used to measure the interoperability among document formats, document templates and individual documents. It provides an effect way for evaluating the results of document interoperability testing which is essential to the assurance of system interoperability.
Abstract: In order to support selection of partners in Computer supported cooperative learning (CSCL), an evaluation model synthetically evaluating user trust and user capability (UT-IOCEM) is proposed. Trust measure model and strategies are discussed, in which a dynamic direct trust measure of individual by someone is realized by increasing the weight of last interaction trust with the rate in sum of squares of all interaction trust, and integrated evaluation of individual trust are also discussed according to the strategies. Following is individual oriented capability evaluation model, which is sum of N dimensional capacity attributes with the corresponding weights. At last, the proposed model UT-IOCEM is validated by a prototype development of UT-IOCEM architecture and used for flexible cooperative action delegation.
Abstract: This paper proposes an efficient statistical method to analyze temperature variations for few hot spots in a 3D thermal analysis. The new method, called EPSN-SOR (and its non-preconditioned version SNSOR), is based on a novel localized relaxation and iterative scheme. The new method can perform statistical analysis on single spot at a time. EPSN-SOR employs evolution and pre-condition techniques for speedup by exploiting topological similarity of two nearby spots. The method further considers spatial correlation in energy and conductance disturbances. Experiments show that EPSNSOR is about three orders of magnitude faster than the Monte-Carlo method with small errors and is about 80X faster than general global SOR (successive over relaxation) method in statistical analysis of 450 hot spots for a test case with 1.3 million nodes (average 0.0553 second per spot) under spatially correlated energy and thermal conductance disturbances on a normal desktop PC.
Abstract: Getting a compact representation of belief spaces is one of the most important issues in a conformant planning task. In this paper, a translation approach to Conformant planning tasks in a Finite-domain representation, noted the CPT-FDR, is studied to translate conformant planning tasks specified in PPDDL formalism into a concise grounded representation that uses finite-domain state variables. It is extended for the semantic of the nondeterministic effects, fluents, axioms, and extended belief states. The FDR-based translation algorithm employs several techniques that can deal with the uncertainty in the initial state and in the non-deterministic operator effects. The experimental results show that the approach can map the conformant planning tasks in the PPDDL into finitedomain representation, and CPT-FDRs can save the memory space and reduce the size of belief states effectively.
Abstract: The multi player pursuit-evasion game is a very important issue in multi-robot system. A hierarchical approach is proposed, then the multi player pursuit evasion game is divided into many single two player game and some assist to execute constringent pursuit. The Fast marching method (FMM) is proposed to solve the Eikonal equation of each pursuit and evader's current movement curve, then to construct the pursuit and evader' “active regions” in the constringent pursuit and forecast their movement. Experiments show that the hierarchical approach based on FMM is more efficient than any other methods in time capability.
Abstract: To enhance the integrity and safety of the flight, the User differential range error (UDRE) must bound the maximum residual error in the service volume with the probability of 99.9%. The most important thing of computing the UDRE is how to find the Worst user location (WUL) in the service volume. Then, a UDRE algorithm using PSO is described to solve this problem. To avoid falling in local optimum, the proposed algorithm generates a group of particles with uniform spacing in the service volume. Experiments were made to compare the performance of the proposed PSO algorithm, the original PSO algorithm and the exhaustive grid search method. The related results show that the proposed PSO algorithm can find the WUL precisely and increase the accuracy of UDRE and this algorithm is suitable for computing UDRE at the master station.
Abstract: The commonly-used image Superresolution (SR) reconstruction methods are computationally expensive and only applicable to “offline processing” cases. To reduce the work-load of SR fusion, three algorithms, i.e., Neighborhood searching nearest interpolation (NSNI), Fixed-distance searching nearest interpolation (FDSNI) and Two-direction linear interpolation (TDLI), are proposed in this paper in the condition of multiple lower resolution images with global random shifts. The simulation results illustrate the effectiveness, speed and quality advantages of the three proposed SR algorithms over classic algorithms.
Abstract: Small scale of labeled samples results in incorrect of computation of mutual information, which may lower the classification accuracy of minimal-redundancymaximal-relevance (mRMR) selective Bayesian classifiers. In order to solve the above problem, a kind of selective Bayesian classifier based on semi-supervised clustering algorithm is proposed. At first, a new semi-supervised Krepresentative clustering algorithm is designed by using the Bayesian posterior probability, which is applied to labeling the unlabeled samples so as to enlarge the scale of labeled samples. Then a novel feature selection criterion is proposed by combining the mRMR and the concept of Markov blanket to automatically determine a reasonably compact subset of features. In addition, a risk-regulation factor is introduced into the feature selection criterion to reduce the risk of mislabeling. At last, a Bayesian classifier is constructed based on the preprocessed samples. Experimental results indicate that the proposed Bayesian classifier can select optimal features to obtain high classification accuracy.
Abstract: The code tracking error model of coherent Code tracking loop (CTL) is modified since existing model cannot accurately describe the characteristic that the code tracking error is determined by multipath and correlator spacing in multipath. The noncoherent DOTproduct (DOT) power mode CTL is established as a comparison with different CTLs in multipath. Besides, the carrier tracking error model is modified based on the established code tracking error models, which indicates that the carrier tracking error is also determined by the multipath and the correlator spacing. The presented error models indicate that the correlation between the code tracking error and carrier tracking error can be decorrelated. Theoretical analysis shows that the noncoherent DOT power mode CTL is essentially coherent when the discriminator output of carrier tracking loop is zero. The influence of multipath and correlator spacing on CTL and carrier tracking loop is analyzed by simulations.
Abstract: In this paper, we propose a speech enhancement algorithm which has the feature of interaction between adaptive beamforming and multi-channel postfiltering. A novel subband feedback controller based on speech presence probability is applied to Generalized Sidelobe Canceller algorithm to obtain a more robust adaptive beamforming in adverse environment and alleviate the problem of signal cancellation. A multi-channel postfiltering is used not only to further suppress diffuse noises and some transient noises, but also to give the speech presence probability information in each subband. Experimental results show that the proposed algorithm achieves considerable improvement on signal preservation of the desired speech in adverse noise environments over the comparative algorithms.
Abstract: The sea surface height recovery methods and error analysis of altimetry using GPS reflected signal are explored in this paper. For the case when the receiver is at low altitude positions, the analytical equation of the sea surface height error is obtained; and for the case when the receiver is at high altitude positions, a novel method which has a concise mathematical formulation and efficient solution is proposed. Moreover, the proposed method can obtain more accurate solution than methods currently available in the literature. Analysis and simulation show that the sea surface height error and all the measurement errors are of the same order of magnitude when the elevation angle is high.
Abstract: In this paper, we try to introduce the idea of total variability used in speaker recognition to language recognition. In language total variability, we propose two new recognition systems, Language-independent total variability recognition system (LITV) and Languagedependent total variability recognition system (LDTV). Our experiments show that language-independent total factor vector includes the language dependent information, what's more, language-dependent total factor vector contains more language dependent information. These two systems LITV and LDTV can achieve performance similar to that obtained with state-of-the-art approaches. Experiment results on 2007 National Institute of Standards and Technology (NIST) Language Recognition Evaluation (LRE) databases show LDTV gains relative improvement in Equal error rate (EER) of 23.2% and in minimum Decision cost value (minDCF) of 14.2% comparing to LITV in 30-second tasks, and we can obtain further improvement by combining these two new systems with state-of-the-art systems. It leads to relative improvement of 21.1% in EER and 23.1% in minDCF comparing with the performance of the combination of the MMI and the GMM-SVM systems.
Abstract: This paper proposes a new model for image denoising and decomposition by non-convex functional minimization. Instead of using the Banach norm as the fidelity term, we use the square of L2 norm of the residual component divided by BV semi-norm as the fidelity term. This non-convex fidelity term has very low value for the texture image and high value for the geometric image, so it is appropriate for image denoising and decomposition. The gradient descent procedure is used to solve the proposed minimization problem, which leads to evolve a new nonlinear integral-differential equation to steady state. The experimental results demonstrate the proposed model not only obtains higher SNR but also makes visual improvements compared with the classical TV and OSV models.
Abstract: On the base of the method of Support vector data description (SVDD), this paper proposes a SVDD method based on maximum distance between two centers of spheres. It applies two hyperspheres to separate two kinds of training instance and maximize distance between two centers of hyperspheres at the same time. Experimental results show that the method is effective, especially for unbalance problem, our method can get better results than all other methods.
Abstract: The convergence of voice and data onto IP based infrastructures demands Quality of service (QoS) mechanisms that can be customized to meet diverse end to end performance requirements. This paper focuses on the improvement of the real-time quality of service by revisiting existing packet queuing and forwarding strategies and investigating its potential benefits with our theoretical insights. The theory is thoroughly validated through extensive numerical simulations which clearly verified RED as a decay rate filter to provide enhanced service Quality of experience (QoE).
Abstract: To increase the efficiency of eavesdropping detection in the “Ping-pong” protocol, an improved secure detection strategy based on W states in “Ping-pong” protocol is presented. TheWstate, the decoy particles, is used to detect eavesdroppers. In the experiment, the method of the entropy theory is introduced, and three detection strategies are compared quantitatively by using the constraint between the information which eavesdropper can obtain and the interference introduced. If the eavesdropper intends to get full information, the detection rate of the original “Ping-pong” protocol is 50%; the second protocol is also 50%; and the presented protocol 58%. The experimental result shows that the efficiency of eavesdropping detection in the presented protocol is higher than the other two, so the detection strategy in the protocol can ensure the “Ping-pong” protocol more secure.
Abstract: Increasing operational complexity becomes the major technical challenge for practically deploying and managing mechanisms to guarantee acceptable real-time SLAs. This paper exploits a cost-effective coordinated approach that combines OSPF-TE with new advances in configuring existing active queue management algorithm. Extensive numerical simulation result suggests that a simple, coordinated approach, in keeping with the Internet paradigm, can achieve increased load for a given VoIP quality level and graceful degradation of Quality of experience (QoE) in presence of network congestion.
Abstract: In order to propose the method of evaluating and improving the real-time performance of EPA (Ethernet for plant automation) industrial Ethernet, the real-time performance of EPA information transmission was studied from theoretical and experimental perspectives. By analyzing information transmission regularity and EPA deterministic scheduling mechanism, several indicators including delivery time, time synchronization accuracy, utilization percentage of configured timeslice, datasending time offset accuracy and throughput RTE that can be used to specify the real-time performance of EPA information transmission were presented and investigated. On this basis, the test principles and test methods of the indicators were studied. Furthermore, an experiment platform was developed to test the real-time indicators of EPA information transmission. According to the analysis and the experiment, some methods to improve the real-time performance of EPA information transmission were proposed.
Abstract: Multipath routing protocols can effectively provide delay guarantees for delay-sensitive traffic such as live stream video over Wireless mesh networks (WMN). However, most existing multipath routing protocols rarely optimize the traffic over the multiple paths to promote network performance. In this paper, we propose a novel multipath routing protocol using convex optimization methods to formulate the problem and derive a multipath protocol through an optimization decomposition. Simulation results show that the proposed protocol dynamically and intelligently decomposes the contending traffic over multiple paths and hence efficiently supports delay-sensitive traffic.
Abstract: In light of an unsatisfactory solution to QoS provision for real-time traffic on the Internet, Flowaware networking (FAN) was proposed as an alternative to existing Quality of service (QoS) architectures. Compared with IntServ and DiffServ, it provides fine-grained services and proves to be scalable. In spite of rich literatures investigating the performance issue of FAN in wired networks, little work has been done in wireless network environment, which has been gaining great popularity in recent years. In this paper, we took the first step to design a framework for the performance evaluation of Wired-cum-wireless networks over FAN (WoFAN) using Stochastic network calculus (SNC). Different from most studies which focused on the deterministic service guarantees provided by FAN, this paper shed light on the quantified stochastic analysis of its achievable level of performance. Numerical analyses are conducted to prove the effectiveness of the proposed framework.
Abstract: The advent of wireless networks andWorld Wide Web enables more and more users to access Internet services by means of mobile devices through wireless links. However, most mobile devices are batter-powered and thus energy efficiency is a critical factor. The Wireless network interface card (WNIC) is a significant contributor to the total energy consumption on many mobile devices. Therefore, we propose a novel WNIC power-saving scheme for Web browsing applications, the scheme tracks network connections and get hints of when transferring WNIC into low power mode to save energy. In the request/response cycle of Web browsing, the network is in idle state during the waiting-for-server period and the connection-inactive period, during those periodsWNIC could be transferred to power-save mode. The length of waiting-for-server period is predicted based on round trip time of the connection, while connection-inactive period is identified with hints got from HTTP and TCP protocol. Real-world experiments show that our scheme achieves a good balance for the competing goals of application performance and energy conservation.
Abstract: To solve the question that players will deviate from secret sharing protocol for their own benefits, we apply game theory to the protocol. It is the first time to use the model of dynamic game of incomplete information, by controlling the probability parameter, the designated strategies will be a computational perfect k-resilient Nash equilibrium. Using this protocol, to achieve fairness, we don't have to assume the number of malicious players is no more than half of the total number which is important in traditional protocols, our protocol can prevent deviating behaviors effectively to achieve extreme fairness.
Abstract: We investigate the problem of precoding optimization in an Amplify-and-forward (AF) Multipleinput-multiple-output (MIMO) relay system. Most reported works on this problem focus chiefly on the design of relay precoder without simultaneously optimizing the direct link. In this paper, we propose a method for joint source/relay precoder design, taking both direct and relay links into account. Our design is based on maximizing the Mutual information (MI) under limited transmission power constraints at the source and relay, respectively. We first formulate a constrained optimization problem before relaxing the original cost function for tractability and derive a MI lower bound which asymptotically approaches the exact expression of MI in an iterative fashion. In contrast to previous strategies, we then prove that the optimal structure of the source and relay precoders jointly convert the MIMO relay channel into a bank of Single-input-singleoutput (SISO) relay channels without having to assume a beamforming structure to simplify the derivation. Specifically, the linear precoding design problem degenerates into power loading among multiple SISO relay channels. Applying standard Lagrange technique results in a scalar convex optimization which can be readily solved by iterative water filling. Numerical examples demonstrate that the proposed scheme, either exploiting partial or full Channel state information (CSI), significantly outperforms the existing methods.
Abstract: N-frequency Transmission line transformers (TLTs) with N-section for two arbitrary complex frequency-dependent impedances is optimally designed based on an ideal transmission-line model. Optimum design equations of the proposed transmission line transformer are derived and solved by using Particle swarm algorithm (PSO) and Levenberg-Marquardt algorithm( LMA). Several numerical examples are given to verify the validity of the design.
Abstract: For defects of image filtering algorithm based on the modified Information entropy (IIE), a novel fusion image filtering algorithm based on edge point judgment by fusion entropy is proposed in this paper. IIE algorithm only fuses the decision of edge point and it still uses single method (linear or nonlinear filters) for image pixel. So it is not a real fusion filtering algorithm. Therefore, the proposed algorithm not only applies fusion theory to edge point judgment, but also uses fusion theory for fusion filtering which can improve the performance of filtering. The experimental results indicate that the proposed algorithm is effective.
Abstract: A novel sine waveguide slow-wave structure was proposed for sheet electron beam traveling-wave tube. Its dispersion and radio frequency transmission characteristics were calculated in W-band. The electromagnetic wave amplification capability based on interaction between sheet electron beam and slow wave was gained through particle-in-cell method. The investigation results show that, the sheet electron beam sine waveguide travelingwave tube can produce more than 100 W from 85GHz to 102GHz with a voltage of 20.3kV and a beam current of 150mA.
Abstract: This paper presents a novel 2.4GHz Chebyshev bandpass filter that exhibits low insertion loss in the passband and high harmonic suppression in ultra-wide stopband. Asymmetric compact microstrip resonant cells combined with a shunt open-circuited microstrip line are employed as admittance inverters instead of the traditional quarter-wavelength lines to suppress the second, third and fourth harmonics in the upper bands. Folded lines are used to realize the half-wavelength shunt open-circuited stubs to achieve a size reduction and performance improvement. By introducing a new optimization strategy, the filter can be efficiently optimized by separately optimizing the resonant cells and folded lines. Measured results show very good performance with a low insertion loss of 0.69dB and return loss of 30.5dB at the center frequency with 0.01dB equal-ripple bandwidth of 5%. The ultra-wide stopband is from 3.37GHz to 9.88GHz, where harmonics up to the fourth order are highly suppressed.
Abstract: A joint range-velocity closed loop tracking filter, where velocity measured value is adopted in the range tracking filter to modify the velocity and acceleration equations, is proposed. The tracking loop use the proposed filter can not only track the range and velocity simultaneously, but also improve the range tracking precision significantly. The experimental results show that the tracking errors about range thermal noise in the proposed loop is lower than that in the traditional loop over 2.2dB. Moreover, with the increase of the filter parameter, the tracking performance of our schemes are improved accordingly.
Abstract: In order to correct ionosphere error on the GNSS signals, this paper proposes an algorithmto estimate the ionosphere corrections on the basis of the observability analysis. The distributed Kalman filter consisting of some parallel sub-filters is utilized and the ionosphere corrections are obtained from weighting the results of these sub-filters. However, the performance of some sub-filters is degraded by the sub-filters with poor observability. To evaluate the observability of these sub-filters, the Ionosphere observable factor (IOF) is defined and calculated by the eigenvalue decomposition. Moreover, to mitigate the adverse effect from the poor observability and boost the accuracy of estimates, an observability-based weighting strategy is designed and adopted. The simulation result demonstrates that the a significant improvement on the accuracy is gained by the observability-based weighting strategy.
Abstract: The optimal radar waveform for target information extraction had been derived to be the waterfilling solution using maximum Mutual information (MI) criterion, however, the optimal waveform is not a constantenvelop signal. In order to make full use of the system radiated power, a phase-iterative method to design the constant-envelop phase-modulated waveform is proposed. By minimizing the Euclidean distance between the designed waveform and the optimal one, the proposed method can achieve a small MI loss. The results of extensive simulations demonstrate that the MI loss of our approach can be less than 0.5nat when the signal duration is greater than 1μs.