摘要:Video steganalysis has become an important research direction of the information hiding.How to search for motion vector(MV)hidden traces in the compressed stream with various video formats and design an effective steganalysis algorithm are hotspots in current studying.Movement in the video can be divided into two groups:the imaging equipment movement and the movement of the subject.Due to the block-matching motion estimation in H.264,strong correlations among adjacent macroblock's MV are existed.However,MV-based information hiding will cause anomalies in the correlations between MV.This paper proposes a method of correlation measurement based on anomalous co-occurrence frequency between MV,and on this basis,combined with H.264 coding technology related to MV,designs a steganalysis algorithm for H.264 video.Experimental results demonstrate that the proposed method has the advantages of high speed,and can be used to effectively detect MV-based steganography in the H.264 video.
摘要:Hash resource required for RFID-tags chips,a secure and lightweight cipher implementation is proposed to solve the threat posed by differential power analysis(DPA).The proposed implementation is based on Piccolo cipher,in combination with the recently proposed masking method named threshold.Boolean equation rearrangement and improved exhaustive search method was respectively applied to S-Box and inverted S-Box to get their optimal area costs for threshold(3,3)share.A method based on latch is proposed to resist the latent glitch threat.Based on Chartered 0.18μm and 100 kHz operating frequency for RFID applications,the proposed cipher implementation costs as low as 2155 gate equivalents and consumes 2.60μA average current.The security of the implementation is evaluated based on real power traces from an FPGA platform.DPA attack results show the proposed implementation with threshold countermeasure is suitable for secure low-cost passive RFID-tag ICs.
关键词:RFID-tags IC;security;Piccolo;differential power analysis(DPA);threshold
摘要:The expressiveness of first-order queries makes first-order query rewriting based on database repair has more practical value on solving CQA problem over inconsistent database compared with other methods.However,the execution efficiency of consistent query rewriting generated from the existing methods might be unsatisfactory.How to effectively promote the performance of consistent query in the data integration environment is studied principally in this paper.Facing the same conjunctive query class Cforest,the consistent query rewriting algorithm ConsRewrite_OR based on OR-database integration schema is presented.Both query cost analysis based on the linear work metric and TPC-H simulated experiments on Oracle 10g adequately indicate that the consistent query rewriting produced by ConsRewrite_OR can obtain more optimized execution efficiency on integration databases,compared with Fuxman's consistent query rewriting.
摘要:Robust principal component analysis(RPCA)is a very effective method to recover both the low-rank and sparse components.This paper extends RPCA to the case of tensor and proposes a framework of multilinear robust principal component analysis(MRPCA).First,it establishes the model of MRPCA which minimizes a weighted combination of the tensor nuclear norm and l1 norm.Then,it employs the augmented Lagrange multipliers algorithm to solve the above nuclear norm optimization problem.Experimental results demonstrate that MRPCA is more robust than RPCA for the data with multilinear structure.
关键词:multilinear robust principal component analysis;robust principal component analysis;low-rank;nuclear norm minimization;augmented Lagrange multipliers
摘要:At present,the conventional image recognition algorithm cannot be well applied to mobile platforms due to their large computational cost and high storage requirements.In this paper,a novel image recognition algorithm for smartphone is proposed:Firstly,BRISK detection algorithm and low-byte FREAK descriptor are used to extract image feature and represent image respectively,which solves the problems of long detection time and large memory footprint;Secondly,BRISK features are improved to accomplish the distinction between the similar structure features by adding the gravity information of smartphone to the image feature;Finally,multi-level index of descriptors have been established to achieve fast finding on similar descriptors which solve the matching problem.The experimental results show that the proposed algorithm can be effectively run on the resource-constrained general smartphone to achieve scene recognition in real-time.
摘要:One of the major challenges of video coding is how to attain the best coding performance under different constraints of computational complexity with various hardware.The computational scalability of coding algorithms is particularly important on this occasion.On the other hand,more and more complicated coding structures are employed to repeatedly applied motion estimation on each block.And the algorithms of motion estimation exerts significant influence on the overall performance of video coding.Therefore,we proposed an optimized motion estimation algorithm with scalable complexity.This algorithm can automatically adapt to different video contents and optimally allocate the computing resources by setting appropriate distortion thresholds.These thresholds are the minimal predicted distortion of each macroblocks plus a unified offset.And all of these thresholds can be computed easily with negligible costs.According to our experiments,our proposed algorithm not only provides scalable complexity,but also achieves better coding performance at the same computational expenses.
摘要:Mobile Augmented Reality system presents surroundings in a way of virtual-real fusion for location-based services.This paper employs C/S architecture for vision-based mobile AR system,which solves the problem of processing speed and storage space on mobile phone.Vocabulary tree algorithm which is based on distributed clustering algorithm is used for city-scale scene recognition on the server.BRISK feature and optical flow algorithm is used in combination for visual tracking on mobile devices.Experimental results on UKbench and SUN397 show that the proposed distributed clustering algorithm can quickly build vocabulary tree and solve the limitations of memory.Mobile terminals keep real-time tracking through extraction of BRISK feature and design of parallel pipeline.
摘要:Dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) suffer artifact by patient motion during the imaging procedure.It is necessary to correct patient motion by deformation registration for DCE-MRI.We present a MAP framework of joint estimation of the enhancement field sequence and the deformation field,which can make use of the prior knowledge of enhancement in space-time.The enhancement field sequence and the deformation field are modeled by the MRF models with the smoothness constraints,and are estimated by a two-step alternate optimization algorithm.In the alternate optimization procedure,the estimated enhancement field sequence is used to de-enhance the post-contrast images.Thus,the pre-contrast and post-contrast images can be considered as the same modal image for registration and it make the estimation of deformation field easier.The experimental results show that the proposed method can estimate the enhancement field sequence accurately and provides high registration accuracy.
关键词:breast tumor;DCE-MRI;image registration;markov random field model
摘要:MARTE is utilized for embedded software modeling.However,it could not provide powerful methods for rigorously analyzing the correctness of software systems.This paper introduces the PTA-OZ,which combines the specification language Object-Z with the probabilistic processes PTA.It can provide accurate descriptions for both static and dynamic semantics of embedded software models.Additionally,we propose the model transformation rules for converting MARTE models to PTA-OZ models.We verify the semantics preservation in model transformation,proving that the model transformation rules can keep both the structural and behavioral consistency of software.A practical case is demonstrated throughout the processes of software modeling and property verification.
摘要:DE algorithm is simple and efficient,but for complex problems also exist the problem of low efficiency of convergence.In order to improve the global exploration ability and convergence precision,this paper proposes a novel elite local learning dynamic differential evolution algorithm.Firstly the history elites are preserved in the elite pool,and then the elites in the pool conduct local learning by sine functions,finally dynamic DE model is used to effectively improve the speed of convergence,and the convergence of the algorithm is proved in theory.Algorithm has been tested on 20 benchmark functions including unimodal functions and multimodal functions and shift functions,experiments result verified the effectiveness and applicability,and the new algorithm can maintain higher convergence speed while maintaining better convergence accuracy.Comparison with the state-of-the-art DE in statistical analysis proves that the algorithm is a kind of new competitive algorithm.
关键词:differential evolution;elite pool;elite local learning;dynamic differential evolution
摘要:Heuristic algorithms based on state search are usually used to solve scheduling problems of Petri nets.Suffering from the state space explosion,their executions are often not efficient for large-scale problems.This paper proposes a time Petri net with ant colony mechanism to address this problem.We introduce pheromones and foraging rules of ants into transitions and evolution rules of a time Petri net.The proposed model can work out scheduling problems during the run of a Petri net without building reachability tree.We use this model to describe and solve an interval job shop scheduling problem.The experimental results further verify the effectiveness of the proposed model.
关键词:scheduling optimization;ant colony optimization;time Petri net;flexible manufacturing system
摘要:In interactive genetic algorithms with a large population,the increase of evaluated individual aggravates user fatigue,which restricts the applications of these algorithms.In this study,a method of selecting individuals using the elite set was presented.The elite set is first formed based on individuals with high user's evaluations;and then,individual categories similar with the elite set are selected to perform genetic operations with neither user's evaluations nor fitness estimations;finally,the elite set is updated according to evolutionary stages and individuals' contributions to the elite set.The proposed algorithm was applied to a curtain evolutionary design system,and compared with existing typical ones.The experimental results confirmed that the proposed algorithm has advantages in alleviating user fatigue while improving its efficiency in exploration.
关键词:interactive genetic algorithm;qualitative index optimization;elite set;individual;entropy
摘要:On the basis of analyzing the rough set theory and immune computational theory,a hybrid attribute reduction algorithm is proposed in this paper.In order to enhance the diversity and stability of the antibodies group,the attribute kernel parameter is taken into antibodies coding.Then the population is vaccinated stochastically through a bacterin extraction algorithm.The approximate quality is taken as the affinity function objection,which is in order to reduce computational complexity of this algorithm.A niche immune sharing mechanism is introduced in the optimization process,which can dynamically adjust the affinity of antibodies and improve the local searching ability of the population.Through all these operators to prompt the convergence speed and kept the balance of global and local optimization.The experiments of the rolling bearing fault diagnosis and UCI data sets reduction have illustrated that the algorithm has outstanding advantages in precision and efficiency.
摘要:The present paper proposes the definitions of state description and existing state for a formula with respect to a maximal consistent theory.Some important properties of consistent theory are obtained by using state description as the tool.Then,it is proved that maximal consistent theory is just the theoretical closure of consistent exist state of all simple disjunctive formulas and conjunctive ones.
摘要:A novel user interest mining method is proposed.Firstly,the items of visiting behaviors are retrieved from user's search engine log,and individual user interests with every webpage are described through the concepts of common ontology.Then,a method for computing the score of interest is proposed.According to the scores,a user's interest list can be judged as different interest patterns,which can be used to find the user's short term interests.After that,a user's interest model is built with concept collection extracted from ontology.At last,based on incremental accumulation of short term interests,long term interest collection can be calculated.The whole procedure avoids the problem of using similarity computation and document clustering to merge concepts in existing interest mining methods.This paper explores a new way of thinking.And as the experiment shows,the proposed method provides a more concrete description of user interest model and obtains an optimized concepts merging result.
摘要:In order to solve the issue of access selection in heterogeneous wireless networks,a new algorithm based on chaos genetic algorithm is proposed,of which the issue is converted into a multiple attribute optimization problem,then the chaos genetic algorithm is used to solve the global optimization problem.Firstly,a hyperchaotic system is used to generate the initial population and the chaos disturbance vector to improve the genetic algorithm,then the network selection evaluation index weight can be obtained by solving the fitness function.The simulation results show that the newly arrived users can be distributed more evenly by using the algorithm.Meanwhile,the network blocking probability and the disparity of the blocking probability of different candidate networks can be reduced effectively,so that the network load is balanced.
摘要:Continuous loop averaging deconvolution (CLAD) is a newly developed method to recover the transient auditory evoked potential (AEP) from responses evoked by high stimulus rate sequences.This method brings forward a challenge in the generation of an appropriate stimulus sequence which will affect the deconvolution performance.We proposed a variant of differential evolution (DE) algorithm,namely solution-space contraction differential evolution (scDE) algorithm to optimize the sequence associated with an objective function defined in terms of the jitter ratio (JR) and the constraint condition in frequency domain.A dynamic scaling factor F in the mutation process was formulated to guarantee the gradual reduction of the searching-space.The scDE algorithm can be efficient in generating required sequences under various conditions with lower JR.This study is thus significant in promoting the application of CLAD method in basic and clinical research.
摘要:A Ground Penetrating Radar (GPR) UWB transmitting/receiving antenna system mounted on a vehicle for the exploration of the subsurface is investigated in this paper.A dipole antenna with two half-elliptical-shape arms which are loaded by continuous resistive elements can broaden the bandwidth and improve the impulse radiation.The influence of the height of backed cavity to the antenna electrical performance is analyzed.Taking fully into account the antenna radiation efficiency and the waveform fidelity,a backed cavity with height of 22mm is adopted.The antenna system is manufactured and mounted on a vehicle.Then it is measured in an anechoic chamber.The results of simulations and measurements are compared together and good agreements are achieved.The results show that the UWB transmitting/receiving antenna system is suitable for vehicle-mounted GPR applications.
摘要:In cloud computing system,most of the existing cloud resource management methods are developed for the resource providers,while few for the users.This paper proposed a satisfactory marriage method between the cloud tasks and resources to meet the preferences of the resource user as well as the provider fairly.The user's requirement of cloud resources and the cloud provider's resource management strategies are extracted as their preferences to the other side.Then,the both-side satisfactions of a resource allocation are calculated.The cloud system resource management system is modeled as a weighted bipartite graph.The existence of the satisfactory marriage scheme is proved and an algorithm is developed based on graph theory to get it.The experiment results demonstrate that our method is valid.
关键词:cloud computing;resource management;two-sides preference;comprehensive satisfaction degree;satisfactory marriage model
摘要:Bring a strong security property,resistance to ephemeral key reveal attack which is needed in the authenticated key agreement (AKA) protocols,to the AKA protocols using smart card and password.A strongly secure two-party AKA protocol using smart card and a strongly secure three-party AKA protocol using smart card were proposed respectively.Meanwhile,the first security model for AKA protocols using smart card and password which includes the ephemeral key reveal attack was proposed.The security proof of the proposed protocols was given in the new security model.The reason why the property of ephemeral key reveal attack can not be achieved in AKA protocols using only password was also given in this paper.
关键词:authenticated key agreement protocol;ephemeral key reveal attack;smart card and password;security model
摘要:Sole threshold based monitoring algorithms bring down the accuracy of alarm,hence researching on dual threshold based monitoring technique,that is,constraint violation monitoring with probability guarantee has significant meanings.Firstly,according to the semantics of probability threshold on the monitoring result,dual threshold monitoring of sensor node was investigated.Secondly,a tight upper bound of the probability of the sensing data larger than monitoring threshold was given,and dual threshold based distributed monitoring algorithm was proposed.Thirdly,based on the given accuracy requirement,a mathematical method to determine an optimal sample was provided.A sampling based approximate cluster monitoring algorithm was proposed.The theoretical analysis and performance evaluation demonstrate the efficiency of the proposed algorithms.
摘要:To reduce the impact of distance information errors,an improved multilateration algorithm named KC-Multilateration is proposed.It is explored that K-means clustering methodology is employed to wireless sensor network localization schemes.By cluster analysis,KC-Multilateration algorithm can figure out the distance data which are far more beyond their true value,and then removes those data from the measured distance data.Then multilateration is adopted with the rest distance data,obtaining the final results.Simulation experiments indicate that the proposed KC-Multilateration can reduce location errors effectively and has more stable location results in a variety of error environment comparing with the original multilateration algorithm.Further experiments based on RSSI indicate that the improved algorithm has smaller location errors and stronger performance of fault tolerance without adding any costs of communication,which verifies effectiveness and practicality of KC-Multilateration.
摘要:We formally prove the range limit of topology observed by a BGP monitor based on the "valley free" model.By comparing with the actual BGP data,we found that for more than 90% of monitors,the observed P-P links are less than 30% of its range limit.By analyzing the relationships between the characteristics of observed ASes and the completeness of observed topology for those ASes,we found that the completeness of the topology is impacted by the distance to the BGP monitor,its adjacency,and other factors.
摘要:WLAN signal is not only hoped to be strong enough in work area,but also to be attenuated rapidly in no-work area,so content of WLAN signal cannot be intercepted illegally.This paper proposes a novelty solution of a smart antenna based WLAN Wireless Signal Coverage Control.Firstly deploying three kinds of sensor,including Inter Sensor,Edge Sensor,Outer Sensor.Initializing object values of weight value and received RSSI.Then through the Particle Swarm Optimizer,combinations of beam and Tx powers of WLAN APs with smart antenna will be determined,so that the wireless signals are the strongest in inter work area,are acceptable on the boundary of work area and the weakest in no work area.Simulations and experiments over a prototype demonstrate that the solution achieved significantly the destination of WLAN wireless signal coverage control.
关键词:WLAN;wireless signal coverage control;sensor;smart antenna
摘要:In order search a more robustness and convenient count method which shows a new feature descriptor algorithm is proposed in this paper.It analyzes and explains this way could quickly and accurately to describe local structure features based on inherit a higher quality of SIFT and MSE.In addition,it makes use of a changed locally linear embedding approach to process data so that it could reduce dimension.Experiment has drawn the conclusion that not only it could obtain more and betters the proposed descriptor but also the count speed could faster than SIFT for the image with zoom,rotation,blurring and illumination varying.This algorithm is suitable for searching the images which has structured features,when it exits multiple of varying.
关键词:multiscale entropy;local aggregation features;locally linear embedding
摘要:The naturalness and auditory quality of the wideband audio is generally degraded due to the limitation of transmission bandwidth,so this paper presents a kind of audio bandwidth extension method from the wideband to super wideband based on grey Verhulst model.Grey Verhulst model was utilized for estimating the envelope of high-frequency spectrum,according to the evolution tendency of the spectral envelope series of audio signals.In addition,nearest-neighbor matching was utilized to predict the fine structure of high-frequency components.At last,the proposed method can effectively recover the high-frequency components in the frequency range 7-14 kHz through the envelope adjustment of high-frequency spectrum.Subjective and objective testing results indicate that the proposed method can improve the auditory quality of the wideband audio and outperforms the conventional method of audio bandwidth extension based on Gaussian mixture model.
摘要:12.5 Gb/s half-rate clock and data recovery (CDR) circuit and 1:2 demultiplexer(DEMUX) applied to a serializer/deserializer (SerDes) receiver were designed and implemented in 0.18μm CMOS process.The CDR and DEMUX provided 6.25 GHz clock and 6.25 Gb/s data for the SerDes receiver.The circuit incorporated a Bang-bang phase/frequency detector (PFD),a four-stage ring voltage-controlled oscillator (VCO),a V/I (Voltage-to-Current) converter,a low pass filter (LPF),a 1:2 DEMUX and so on.A new half-rate PFD which sampled the clock by data was adopted,and this PFD could increase operating rate up to 12.5 Gb/s.The tested results show that at the supply voltage of 1.8V,the tune range of VCO is about 1GHz at 6.25GHz centre frequency.When the data rate of the input pseudorandom is 12Gb/s and sequence length is 231-1,the recovered 6GHz clock has a root mean square (RMS) jitter of 1.9ps and a peak-to-peak jitter of 9.12ps.At the same time,the total chip can work well and the eye diagram of output data is clear.The die size is 0.476·0.538 mm2 and the core power consumption is only 150mW excluding the output buffers.
关键词:serializer/deserializer (SerDes);clock and data recovery (CDR);ase/frequency detector (PFD);voltage-controlled oscillator (VCO)
摘要:It proposed a data dissemination mechanism with network coding based on ant colony algorithm (DDM-NC),used in multi-source multi-sink delay tolerant network (DTN).Based on the publish/subscribe mechanism,the topic data were computed and transferred using network coding method,which can take full advantage of the network capacity to multicast,to improve security and efficiency of the transfer process.At the same time,for working out the problem of large data redundancy and vast invalid deliveries in the packages broadcast process,a package routing scheme based on ant colony algorithm was brought up,to guide the packages deliver to destinations,and cut down the data redundancy and deliver latency.The simulation results show that,compared with the traditional DTN epidemic routing and random network coding method,the DDM-NC method can get better data delivery performance.
摘要:Through the analysis of some mobile malwares,we found that malware is similar with its original application in semantics of the program,and the similarity is different with the similarity between other members of the family.Based on this fact,by means of hierarchical clustering technology for the function call graph,we propose a program based on family relationships to detect the malicious mobile applications and build a system named as "NeighborWatcher".Experimental results show that when each family contains four or more members,the detection rate of Piggybacked application can reach 92.86%.
关键词:piggybacked application;call function graph;family clustering;mobile security
摘要:Trivium is a stream cipher and has successfully been chosen as one of seven finalists by European eSTREAM project.In this paper,a differential cryptanalysis of Trivium based on automatic deduction is presented.This new technique enables the attacker to obtain differential characteristics on arbitrary-round Trivium.The technique is applied to 288-round Trivium,which results in an efficient distinguishing attack.Our attack only requires 226 chosen IVs with a distinguishing advantage of 0.999665.The result is much better than the existing single linear cryptanalysis and Linear Cryptanalysis with Multiple Approximations on 288-round Trivium.We also apply the technique to more-round Trivium and the modified Trivium proposed by Turan and Kara.The results show that Trivium reduced to no more than 359(out of 1152)initialization rounds is weak against differential cryptanalysis,and the modified Trivium is better against differential cryptanalysis than the original Trivium.
摘要:In order to improve the convergence and distribution of Many-Objective Evolutionary Algorithms (MOEAs),this paper proposes a Many-Objective Fuzzy Evolutionary Algorithm (MFEA) which is based on fuzzy dominance.On the model of algorithms based on Pareto-dominance,we improve the environmental selection using fuzzy logic.We present a new dominance strategy based on fuzzy membership.Then,we propose a new estimation method of crowding distance which incorporates Harmonic-distance,k-neighborhood method and niche technique.Finally,according to the characteristics of MOPs and the idea of α-cut set,we design a new environmental selection strategy which is made up of two truncations.The proposed algorithm is compared to 5 state-of-the-art MOEAs on benchmark test problems.Simulation results show that α-MFEA has obvious advantages than other algorithms because MFEA could ensure good convergence while has uniform distribution,especially,applied to solving high-dimensional MOPs.
摘要:As the feature size of integrated circuits shrinks,soft errors become the key factor influencing circuit reliability.Calculation of signal probability of logic circuits affected by soft errors can assist us in evaluating the circuit reliability.The reconvergent fanouts in circuits increase computational complexity of signal probability.A reliability calculation method of logic circuits is proposed in this paper.The method uses the probability formula and polynomial arithmetic,then the fanout source variables which trigger signal correlation are reduced.Based on this,the circuit reliability can be evaluated by the output signal probability.Experimental results on LGSynth91,74X series and ISCAS85 benchmark circuits show that our method is accurate and efficient.
关键词:soft error;signal correlation;fanout reconvergent;reduce order;conditional probability