HUANG Dong-ping, LIU Duo, WANG Dao-shun, DAI Yi-qi
Vol. 34, Issue 11, Pages: 1937-1940(2006)
摘要:A verifiable threshold multi-secret sharing scheme is proposed in this paper.As the secret can be recovered with the shadows provided by participants and it is computationally difficult to get the sub-keys from the shadows,the sub-keys can be reused to share the multi-secret in this scheme.By verifying the information published by the dealer as well as the shadows of sub-keys provided by participants,this scheme can prevent both dealer and participant from cheating.The security of this scheme is the same as that of RSA cryptosystem and Shamir’s (k,n)-threshold scheme.Two kinds of cheating methods against threshold multi-secret sharing scheme are also proposed,which can threaten the security of previous schemes more or less.But the scheme proposed in this paper provides efficient solutions against these cheatings and achieves the same computational security with a better performance compared with the previous schemes.
摘要:In accordance with the increasing requirements for video transmission over unreliable channels,the issue of error resilient video coding technology has been paid enough attention.Based on ZIG-ZAG interleaved flexible macroblock ordering,a novel algorithm called ZFMO is proposed in this paper.Different with the traditional raster scan order,ZFMO achieves a good tradeoff between coding efficiency and error resilience by rearranging macroblocks in ZIG-ZAG scanning order and partitioning the macroblocks into different slice groups in an interleaved way.Experimental results show that ZFMO outperforms H.264/AVC recommended algorithm in low packet loss environments.The using of ZFMO influences intra frame prediction and motion estimation because of the macroblocks rearrangements,so the coding mode selection must be reconsidered.The rate distortion optimization based ZFMO decreases the redundancy bitrate while maintain recovery video quality and error resilience performance.Simulation results based on H.264/AVC reference software JM9 suggest that the performance of rate-distortion optimized ZFMO can be further improved.
关键词:error resilient video coding;flexible macroblock ordering;rate distortion optimization;H.264/AVC
摘要:Portscan is used to figure out whether the target system’s ports are open by trying to access these ports.It is usually the fist step of a sequence of intrusion actions.Portscan detection is an indispensable part of an intrusion detection system.However,there are only a few portscan detection methods nowadays.Moreover,they are not very accurate.In order to improve the accuracy of portscan detection,the data produced by two portscan detection methods is fused using Dempster-Shafer theory of evidence.One method is the ports distribution based portscan detection,which is very simple and has a pretty high detection ratio.The other is the sequential hypothesis testing based detection method,which sufficiently exploits the portscan’s essential character.The experiment shows that the portscan detection method based on Dempster-Shafer theory of evidence is far more accurate than the one base on ports distribution or sequential hypothesis testing.
关键词:portscan detection;intrusion detection;Dempster-Shafer theory of evidence;data fusion
摘要:Data replication is a general mechanism to improve performance and availability for data grid applications.Distributing data replica reasonably in large scale data grid systems can decrease communication cost and improve its performances.This paper proposes a new Calking Dynamic Replication Distribution Algorithm (CDRDA) considering the communication cost among storage resource brokers and the characteristic of data accessing.By the algorithm,data replica will be distributed as a multi-level virtual cache and saved at the node in which applications can acquires the average lower communication cost.At last,the performance of the algorithm and its applications are also introduced.
关键词:data grid;data replica;cost of communication;calking
摘要:Fuzzy Petri Net (FPN) is an extension of Petri Net (PN).Previous research focuses on the representation and reasoning of the fuzzy rules with FPN.However,fuzzy production rules with negative literals can not be expressed and reasoned effectively in traditional methods.The inherent meanings of the negative literals in fuzzy production rules are discussed in this paper,while negations of the input proposition can be represented as the negative impact on the rules,and negations of the consequents can be treated as that the rule blocks the consequents.Based on it,Consistent Fuzzy Petri Net (CFPN) model is proposed,which is more suitable to represent the fuzzy logic programs with negations.Thresholds are also introduced to CFPN model.Finally,a formal reasoning algorithm for the CFPN model is presented and validated.
关键词:fuzzy petri net;fuzzy reasoning;fuzzy production rule;negative proposition
摘要:According to these shortcomings exited in mulitscale associated estimation methods,a new multiscale estimator is proposed by combining wavelet transform and Kalman filtering,in which a new technology of systemic and multiscale transform are employed.The state equation and measure equation described only in temporal domain are rewritten down data-block equations;these data-block equations are characterized by use of multiscale technology in both temporal and frequency domains;the new associated filter is established via using Kalman filter and sequential filter.The new filter not only holds the good characters,such as the real time and recursive,but also possesses the multiscale analysis capability.Computer simulations show that the estimate accuracy of new algorithm is comparable with or little better than that of traditional Kalman filter.
关键词:multiscale associated filter;real-time and recursive;wavelet transform;Kalman filtering
摘要:Breakdown characteristics of the low-frequency dielectric barrier discharge (DBD) at low pressure are investigated experimentally in He,Ne and Ar.The current waveform of this DBD is a series of pulses that is caused by the results of the electron avalanches under the action of the applied field and the quenching effect of the wall-charges field on the avalanches.The experimental investigation and the theoretical analysis on its breakdown characteristics indicate that its breakdown voltage is higher then that calculated by Paschen law and depends respectively on the filled-gas pressure and the distance between the electrodes,instead of their product,due to considering the diffusion loss of the charged particles in the breakdown processes.The ion-induced secondary electron emission coefficient and the mean electron energy at the breakdown moment can also be determined approximately by measuring the breakdown characteristics of the discharge tubes with different distance between the electrodes.The experimental results and the theoretical analysis of this DBD are discussed in this paper.
关键词:dielectric barrier discharge;gas discharge;breakdown characteristics;breakdown voltage
SUN Shu-rong, YIN Hong-xi, WANG Zi-yu, XU An-shi, YANG Shu-wen
Vol. 34, Issue 11, Pages: 1970-1973(2006)
摘要:A new codeword assignment strategy for OCDMA networks is proposed in this paper.Employing a new mathematical model,namely processor-sharing system,static-state throughput ratio and average delay of OCDMA networks are calculated.The results reveal that our strategy performs better than existing strategies.Meanwhile,analysis process is simplified compared with conventional Markov chain model,and corresponding results are indeed consistent with practical situation,thus processor-sharing system is truly applicable to model the OCDMA networks.
摘要:In integrated circuits,the defects associated with photolithography will generally be assumed to take the shape of circular discs in order to perform the efficient estimation of yield and fault analysis.However,real defects exhibit a great variety of shapes.This paper provides a rectangular model of real defects and the model of computing the critical area involved,which take the shape of real defects as well as the features of the layout routing into account.In the aspect of the prediction of fault probability,the simulation results show that the new model may predict the fault probability caused by real defects more accurately than the circular model does.
关键词:real defect;rectangular defect model;critical area model;yield
XIANG Ying, ZHU Ren-jie, Gino Tuccari, ZHANG Xiu-zhong, SHU Feng-chun
Vol. 34, Issue 11, Pages: 1978-1980(2006)
摘要:Digital single side band (SSB) down converter is an important part in software radio receiver as well as in data acquisition terminal of radio astronomy.Because the down converter follows the high speed Analog-Digital sampler directly,the requirement on the capability of processing data is very strict.In this paper,a parallel architecture is presented to implement Wideband digital SSB down converter,which can solve the DSP 'bottle-neck’.
关键词:digital down converter;single side band(SSB);wideband
摘要:To the problem of digital pictures copyright protection,we propose an asymmetric watermarking method based on feature space decomposing.Passing feature space decomposing of digital image,we obtain that the robustness of the approach lies in hiding a watermark in the subspace that is least susceptible to potential modification.In the same time,because we use different matrix operation to embed(secret key)and extract(public key)a watermark,we are able to release all information for the public,except the secret key.Through analysis and constraint the conditions of feature space,the algorithm we proposed can obtain a high detection probability and security,a low false alarm probability.The robustness of the asymmetric watermarking method is demonstrated by a kind of attacks through computer simulation.
关键词:feature space decomposing;asymmetric watermarking;security;robustness
摘要:This paper presents the benchmark test results on various surface potential based charge-sheet models.Compared with the Pao-Sah model,most charge-sheet models predict the inversion charge with different degree errors in different operation regions.In order to model the MOSFET channel current,a semi-empirical channel current equation has to be used in the present surface potential based charge-sheet models,that lead to device physics inconsistency between the channel current expression and channel charge equation,thus result in some channel current errors.These benchmark test results demonstrated that the charge sheet models require some fundamental improvements to maintain the inherent device physics and high accuracy of the Pao-Sah model.
关键词:MOSFETs;device physics;compact modeling;surface potential;charge-sheet model
摘要:The response of SPR sensor includes two components:specific response and non-specific response.The former is due to the capture of analyte molecules by target molecules immobilized on the sensor surface,and the latter is due to the adsorption of non-target molecules to sensor surface and temperature fluctuations or sample variations of composition and concentration.The existence of non-specific response reduces the precision of SPR sensors.In this paper we provide a kind of multi-channel SPR imaging sensor structure to improve the accuracy of SPR sensors.It can discriminate between the specific response and the non-specific response by comparison of different channels,and realize the real-time detection of the biomolecular interactions.This method is demonstrated effective on specific detection by testing for DNA.
ZHAO Bao-hua, ZHANG Wei, LI Jing, QU Yu-gui, ZHANG Ying-tang
Vol. 34, Issue 11, Pages: 1994-1998(2006)
摘要:To bring out a practical energy-efficient protocol of Wireless Sensor Network (WSN),this paper studies the Sensor Node (SN) joining model.First,we refine WSN layers and divide the network layer into Neighbor Discovering (ND) layer,Dynamic Joining Model (DJM) and routing layer.Moreover,we discuss the ND mechanism in detail,reconstruct the DJM with random parameters,provide the corresponding network booting style and verify the whole model above using mathematic analyses,simulations and a realistic test bed.
关键词:wireless sensor network(WSN);neighbor discovering(ND);dynamic joining model(DJM);packet loss percent(PLP);load balance
DAI Xiu-bin, ZHU Hong-qing, SHU Hua-zhong, LUO Li-min
Vol. 34, Issue 11, Pages: 1999-2003(2006)
摘要:Content-adaptive mesh modeling is an efficient method for image representation.In this paper,the minimum cross-entropy algorithm using prior anatomical information,combined with mesh model,was applied to the reconstruction of PET images.In the proposed algorithm,the nodes of mesh model were extracted from a reference image obtained with FBP method;then,the values of the nodes were computed through the minimum cross-entropy algorithm with prior anatomical information.Finally,the whole image was reconstructed by interpolation from the values of the nodes.The performance of the proposed method was tested and compared with other algorithms using a set of simulated data.The effect of the parameters on the result was also studied.
关键词:content-adaptive mesh model;minimum cross-entropy algorithm;reconstruction of PET(positron emission tomography) images
ZHU Yi-hua, SHEN Yi-Jun, WU Xiao-yan, WANG Jia-cai
Vol. 34, Issue 11, Pages: 2004-2007(2006)
摘要:Communication among mobile nodes in a MANET (mobile ad-hoc network) is multi-hop,i.e.,it is performed via a chain of mobile nodes.It suffers from break due to power exhaustion or movement of the mobile nodes staying in the chain.An algorithm for constructing a minimal CDS (Connected Dominating Set) is proposed to solve the problem.Being aware of power and load of mobile nodes,the proposed algorithm can reduce both the probability of communication route breakage resulting from exhaustion of mobile nodes’ power and the delay time of packets to be forwarded in intermediate nodes of a communication route.Additionally,it is significant for designing robust and effective routing strategies in MANETs.
摘要:Permutation Flow shop scheduling (PFSP) is a complex combinatorial optimization problem with strong engineering background,and is of great importance in both theory and application.In this paper,a new PSO based flow shop scheduling algorithm is proposed to generate optimized PFSP schedule.First,limitation of information sharing mechanism in GPSO model is analyzed and new population based information sharing strategy is proposed.PFSP scheduling algorithm based on new information sharing strategy utilizes problem-concerned knowledge to direct its search in the local search procedure.Compared with representative PFSP scheduling algorithms,the proposed algorithm can obtain good balance between quality of schedule and computational cost.Simulation results validate its efficiency.
摘要:Nowadays,NAS and SAN are served as two main network storage systems,both of which have their advantages and disadvantages.With the dramatic increase of the network application,however,a new network storage architecture,Unified Storage Network (USN),in which NAS and SAN are integrated on the base of an IP, has been made.Firstly,by way of a Global Multi-Protocol File System(GMPFS),the USN combines NAS and SAN,achieving a high scalability and a large capacity.Secondly,with an iSCSI module,the USN serves the block I/O and file I/O simultaneously,combining both the advantages of NAS and SAN.Thirdly,through an Autonomic Storage Agency (ASA),the USN provides two types of data channels:a server channel and a high-speed network-attached channel,creating a direct access to the storage device for its client.And in the concerned experiments,the USN has turned out to be an improved network storage system of a great capacity with an ultra-high-throughput for both the file I/O and the block I/O.Compared with the existed NAS and SAN,this new system has proved to be better in terms of its performance,scalability,compatibity,high - availability and price,etc.
GAO Li, YANG Shu-yuan, XIA Jie, WANG Shi-jun, LIANG Jun-li
Vol. 34, Issue 11, Pages: 2018-2023(2006)
摘要:This paper proposes a novel automatic image-segmentation method which is an improvement to the marker-based watershed transform.A new marker-extracted approach is designed to extract the regional minima from the low frequency components of the gradients.The extracted minima constitute the binary marker image.Then the markers are imposed on the original gradients as its minima,and suppress its all intrinsic minima.Finally,the watershed algorithm is applied to the modified gradients to partition the image.Across a variety of image types,it is proven that this new method can obtain meaningful and homogeneous regions with accurate,consecutive and one-pixel wide boundary.Compared with other methods,this system requires fewer computations and simpler parameters and can more efficiently reduce the over-segmentation of the watershed algorithm.
摘要:This paper presents a scalable scheduling scheme,called iterative request-grant-based round-robin (iRGRR),for high-speed crossbars.It overcomes limitation that most traditional scheduling schemes (such as iSLIP,PIM) suffer from poor scalability by simplifying the execution process and reducing the scheduling overhead.iRGRR dramatically reduces the complexity of control messages from an order of O(N) to O(logN),and thus provides fine scalability,and can be used in terabit switches/routers.The simulation results show that iRGRR can achieve nearly the same performance as iSLIP under various traffic models,including uniform and non-uniform destination distributions.In addition,iRGRR provides better fairness and has lower implementation complexity than iSLIP.
GAO Zhen-guo, WANG Ling, ZHAO Yun-long, CAI Shao-bin, LI Xiang
Vol. 34, Issue 11, Pages: 2030-2037(2006)
摘要:Service discovery is expected to be a crucial feature for the usability of mobile ad-hoc networks (MANETs).In this paper,Minimum Dominating Forward Node Set based Service Discovery Protocol (MDFNSSDP) is proposed.MDFNSSDP has the following characteristics.1) MDFNSSDP deliberately reduces the number of nodes in the current node's 2-hop neighbor set that should be covered by the current node.2) Minimum number of forward nodes are selected based on local topology information and history information piggybacking in service request packets (only these forward nodes are responsible for forwarding service request packets).3) The coverage of service request packets is guaranteed.4) Multiple requests can be fulfilled in just one service discovery session.Simulations show that MDFNSSDP is an effective,efficient,and prompt service discovery protocol.
摘要:It is rather important to study deadlocks in the High Level Architecture (HLA) for correctly comprehending time advancing mechanism in HLA standards,designing time management services in Runtime Infrastructure (RTI),and developing HLA/RTI simulations based on logical time.This paper investigates deadlocks in the HLA time management from different aspects,including deadlocks resulted from zero lookahead,from time advance services,and from time management algorithms as well as the resolution of deadlocks.A series of interesting phenomena are also explored,and varieties of cases in which deadlocks must happen or must not happen are also demonstrated.Particularly,the paper brings forward a rather common principle called Water Surface Rule.This rule shows that all programs cannot advance further and all of them are in relatively quiet status when deadlock occurs,which seems to be the calm water surface.In HLA/RTI simulations,all programs in deadlock status have the identical Greatest Available Logical Time (GALT) so that none of them can advance logical time any further.
摘要:Researchers on security database concern more and more the security threats from DBA(database administrator),for its resulting in increasing crime to internet commercial databases which exist in the mode of ASP(Application service provider) recently,and its theoretic and technical hardness.Traditional access control could not provide enough security for the purpose,cryptographic security database becomes a promising scheme favored by its computing complexity of mathematical difficulties.We analyse various and up to date security threats to distributed database application systems,and give an overview for cryptographic access control and encryption database technology.Each class and approach is described and evaluated with its cryptographic principle,algorithm characters and positive and negative performance.We focus on the term of cryptanalysis of the schemes,which is considered the crucial points of the technology.
摘要:With the rapid development of sequencing technologies,molecular data are accumulated with unprecedented pace which enable biologists to reconstruct the tree of life of all the organisms in the world.Moreover,the use of evolutionary trees is a fundamental step in many biological problems,such as multiple sequence alignments,protein structure and function prediction,and drug design.In this paper,we review the development and studying status of phylogeny reconstruction technology.The major limitations in phylogeny reconstruction technologies are analyzed.Then some vital aspects that may be conducted in the future investigations are discussed.
LIU Wen-bin, ZHU Xiang-ou, WANG Xiang-hong, CHEN Li-chun
Vol. 34, Issue 11, Pages: 2053-2057(2006)
摘要:DNA computing is a new natural computing paradigm developed in recent years.Because of its massive parallelism and high information density,DNA computing has become a "hot" field attracted many researchers.In this paper,we review the development of DNA computing in Models,the simulation of Boolean Circuit,large scale Database based on DNAs,and its application in bioinformatics so that readers can have a complete understand of DNA computing.Finally,the future research directions on DNA computing are pointed out.
关键词:DNA computing;genetic algorithm;Boolean circuit;database based on DNA;bioinformatics
摘要:Lose averse utility function is adopted to describe retailers' attitude for treating with the risk,the problem concerning with optimal purchasing strategies of retailers in a two-periods inventory systems with wholesale price contract is investigated under the circumstance of stochastic market information and stochastic demand.It is shown that the optimal inventory strategy in the second period is base-stock policy.Then we investigate the dynamic properties of optimal purchasing quantities in the second period,which are determined in term of quantities in the first period.The analytical expression of optimal purchasing quantities with its' tighter lower bound in the first period is then given.The relationship between optimal purchasing quantities and coefficients of risk averse in every period is analyzed.Finally,conclusions of this paper are validated by some simulation experiments.
摘要:Based on second generation bandelets and combined with multi-level thresholding and Wiener filter,this paper proposes a novel denoising method.As a new multiresolution geometric analysis tool,second generation bandelets can make full use of intrinsic geometric regularity and provide optimal representation adaptively.Multilevel thresholdings satisfy the statistic characteristics of coefficients.Wiener filter can get rid of the texture effects introduced by thresholding denoising.Numerical tests applied on optical images show that the image denoising method avoids visible ringing effects and provides improvements in Peak-to-Signal Ratio.More significance,this technique is used for restraining speckle noise in synthetic aperture radar images successfully in equivalent number of looks and ratio images.
摘要:Said and Pearlman's Set Partitioned Embedded block (SPECK) algorithm is computationally simple and efficient for image compression and progressive transmission,however there is a main drawback of high memory requirement for hardware implementation.We propose a new low memory embedded block coder that keeps those desirable features,and the novelties are twofold.A compact form of the flag maps is introduced to store the significant coefficients and sorting information instead of using lists.A new block depth-finding strategy is developed for searching insignificant sets at sorting stage.Experimental results show that the obtained PSNR values for the decoded images are very close to those of SPECK algorithm and the memory consumption is reduced by 12 times.Compared with another low memory coder (Listless Zerotree Coder,LZC),the PSNR of our algorithm improves at least 1.1dB with a minor increase of memory requirement.This provides an efficient way for hardware implementation of wavelet embedded block coding.
摘要:It is proposed a new blind separation algorithm of ill-condition mixed sources.Observed signals are pre-processed through eliminating redundancy signals so that mixed matrix A is row full rank.Further,we propose a cost function that is a logarithm of a ratio of the covariance of a part sum of recovered signals and the covariance of recovered signals so that optimizing this function transform to solve a generalized eigenvalue problem.Under a loose condition,it is proved that any source signals theoretically satisfying the condition can be separated.The computer simulation shows its outstanding performance on blind source separation approach.
摘要:A fifth-order elliptic low-pass Gm-C filter used in monolithic DAB receiver is presented.The filter integrates an on-chip frequency automatic tuning circuit based on PLL to control the cutoff frequency.The filter has been implemented in 0.35 μ m CMOS process.The measurement results show that the frequency tuning error is within 1%,the dynamic range is 54dB,and the stop-band rejection rate is greater than 40dB.The filter draws 13mA current from a 3.3V power supply.
关键词:continuous-time filter;Gm-C;automatic frequency tuning;CMOS
摘要:The self-organizing management mode in P2P networks leads to a large amount of selfish behaviors among peers.The corresponding solutions to this problem could hardly keep the merits of P2P network simultaneously,such as self-organization or dynamic scalability.Most of the proposed reputation management models or incentive ones use the flooding mechanism to learn historical behaviour information of other peers,which causes excessive incurred packets,and thus limits the dynamic scalability.A novel incentive framework named as ResP2P based on random matching games theory is given in the paper.Peer reputation and its renewal mechanism,along with some essential social norms are introduced in ResP2P model,which stimulates rational peers to maximize their own utility and contribute their free resource.Experiments have verified the validity and efficiency of the incentive mechanism.The relative distributed algorithm can easily be deployed in a P2P networks and satisfied with self-organization and scalability.
摘要:Optical Burst Switching (OBS) is a promising paradigm for the next-generation Internet.In OBS networks,the most straightforward approach to routing a connection is to always choose the same fixed route for a given source-destination pair.One example of such an approach is fixed shortest-path routing.This approach to routing connections is very simple;however,the disadvantage of such an approach is that,if resources (wavelengths) along the path are tied up,it can potentially lead to high blocking probabilities in the dynamic case,or may result in a large number of wavelengths being used in the static case.In this paper,a load-balancing routing mechanism in OBS networks is proposed,which chose a path that balances the assigned wavelengths of every links.Simulations show that,this method is better than shortest-path routing in block probability,and it is nearly as good as shortest-path routing in transmission delay.It can reflect the real-time change of the traffic flow,at the same time;it is good at adapting different topologies.
关键词:all-optical network;optical burst switching(OBS);routing mechanisms;load balance;routing and wavelength assignment (RWA)
摘要:Stereo technology has been widely used in visual navigation systems.Traditional stereo vision navigation systems usually use local method because of the real-time requirement.The accuracy of the results by local method,however,is very low and it can not be used in high performance vision navigation systems.In this paper we propose a novel fast stereo matching algorithm,which utilizes the segmentation information on the epipolar lines to form a tree and get optimized results by using tree dynamic programming procedure.The proposed algorithm can both use much of the relation information between neighboring pixels to achieve a more accurate matching result,and reduce the number of vertices to be solved so that the system becomes very efficient.Experimental results have shown that our algorithm can get good results comparable to the global MRF-based algorithms,while the executing time is much shorter than the latter.Our algorithm is a fast algorithm which can provide reliable 3D information for vision navigation.
QI Ji, LI Xi, HU Nan, ZHOU Xue-hai, GONG Yu-chang, WANG Feng
Vol. 34, Issue 11, Pages: 2094-2098(2006)
摘要:The management of reconfigurable resource is one of the most critical factors concerned deeply with the performance of dynamic reconfigurable systems.This paper presents two associated algorithms,KTVP (Keeping Tasks Vertexes for Placement) and KTVS (Keeping Tasks Vertexes for Scheduling),to maximize the system utilization.These two algorithms arrange the new-coming hardware tasks according to the vertexes' information of the already running tasks.By exploiting a novel encoding scheme,they can validate the feasibility of the scheduling/placement in a fast way,so as to effectively reduce the waste of resource and improve the parallelism of the whole system.The simulation results show that,compared with existent algorithms,KTVP/KTVS can achieve lower task rejection ratio and reduce the execution overhead significantly.
摘要:H.264 is the latest video coding standard developed by ITU-T and ISO/IEC,and its security is becoming a research focus.Several candidate domains are investigated to apply encryption for H.264 and,a new video encryption scheme combining encryption with compression is proposed in this paper.It includes novel simultaneous CAVLC(Context-based Adaptive Variable-Length Coding)and QTC(quantized transform coefficients)encryption and prediction mode,motion vectors scrambling.Theoretical analysis and experimental results show that the scheme is fast,secure,and robust to transmission errors,moreover,it has very limited adverse impact on the compact ratio.
关键词:H.264;simultaneous CAVLC and encryption;video encryption;scrambling
摘要:This paper presents an embedded multi-channel video encoder using single-chip TI DM642 DSP.Not only could it process 4-way video signals real-time,but also it is full-function,high-reliability,small-size cubage,flexible control and low-power consumption.We construct encoder hardware platform by taking full advantage of DM642 DSP,video ports of DM642 can accept multi-channel video signals at the same time,and DSP handle video data parallel with the operation of data transferring using EDMA.By analyzing the hardware architecture and resource of DM642,we presents GMB coding scheme and efficient algorithm for DSP,namely fast motion estimation algorithm by calculating the best position in sub-pixel level,and DSP-based efficient quantization algorithm etc.
关键词:encoder;video;MPEG-4;digital signal processor (DSP);embedded
HE Pei-yu, ZHOU Ji-liu, XIA Xiu-yu, WANG Yong-de, ZHAO Gang
Vol. 34, Issue 11, Pages: 2109-2114(2006)
摘要:A multi-channel acoustic echo suppression model (MCAESM) via second-order blind signal separation (BSS) is proposed.It avoids the inherent non-uniqueness problem in multi-channel acoustic echo cancellation due to strongly cross-correlated acoustic echo-source signals.On the contrary,it makes full use of the cross-correlativity to remove these acoustic echoes.In this model,the multi-channel acoustic echoes from each microphone signal can be effectively separated and suppressed by employing only one additional well-positioned microphone.For the purpose of real-time processing,a new frequency domain BSS algorithm based on second-order statistics,with less computational complexity and stronger robustness,is proposed to verify the MCAESM.Experimental results confirm the effectiveness of the proposed model.
关键词:multi-channel acoustic echo suppression;blind signal separation;second-order statistics;cross-correlativity
CAI Li-jun, LIN Ya-ping, LU Xin-guo, YI Ye-qing, LI Xiao-long
Vol. 34, Issue 11, Pages: 2115-2119(2006)
摘要:Independent component analysis (ICA) is a statistical method applied to gene cluttering.Estimative separation matrix algorithm of ICA uses mainly Random Gradient Algorithm or Natural Gradient Algorithm.And yet these algorithms can only get the partial optimized solution.This paper proposes a new algorithm of gene clustering based on genetic algorithm.The key idea is by using genetic algorithm instead of previous estimative separation matrix algorithms in ICA to classify gene expression data.The former has an advantage of overcoming partial optimized solution.The analysis and experiments support our conclusion that gene clustering based on genetic algorithm has better performance.
摘要:A cellular neural network (CNN) is a large-scale nonlinear analog circuit suitable for real-time signal and image processing.CNN can be used for high-speed parallel computation and is easy to be translated into a VLSI implementation.This paper presents one new approach for shape from shading (SFS) using paralleled hardware annealing CNN that performs optimization algorithm.Some practical results are presented and briefly discussed,which demonstrates the successful operation of the proposed algorithm.This new approach is very affordable to parallelism and analog VLSI implementation,which allowing the SFS solution to be performed in real-time.
关键词:cellular neural networks(CNN);shape from shading (SFS);optimization
摘要:The parallel-distributed fusion system based on soft-decision can fuse the different feature class,and the fusion result is correlative to these output of classifiers.When the vector dimension to different classifiers isn’t equal,whether the dimension difference affects the last result,it’s important to design algorithm and select classifier in the fusion system.To the system of fuzzy integral fusion multi-FasART networks,this paper analyses the causation of dimension difference influence,presents a way to settle the problem.Experimenting with spectrum and texture feature of remote sensing image,the results show that the dimension difference has influences on the fusion algorithm.The fuzzy integral fuse multi-FasART networks that is a typical algorithm of parallel-distributed fusion system based on soft-decision,so algorithms based on this fusion system have the problem of dimension difference.