2010 Vol. 19, No. 4

Display Method:
Actor-Critic Learning Based on AdaptiveImportance Sampling
CHENG Yuhu, FENG Huanting, WANG Xuesong
2010, 19(4): 583-588.
Abstract(745) PDF(1361)
Abstract:
For on-policy Actor-critic (AC) reinforcementlearning, sampling is a time-consuming and expensivework. In order to efficiently reuse previously collectedsamples and to reduce large estimation variance, a kindof off-policy AC learning algorithm based on an Adaptiveimportance sampling (AIS) technique is proposed. TheCritic estimates the value-function using the least squarestemporal difference with eligibility trace and the AIS technique.In order to control the trade-off between bias andvariance of the estimation of policy gradient, a flatteningfactor is introduced to the importance weight in the AIS.The value of the flattening factor can be determined by animportance-weight cross-validation method automaticallyfrom samples and policies. Based on the estimated policygradient from the Critic, the Actor updates the policyparameter so as to obtain an optimal control policy. Simulationresults concerning a queuing problem illustrate thatthe AC learning based on AIS not only has good and stablelearning performance but also has quick convergencespeed.
A Routing Protocol Based on Ant ColonyAlgorithm for Wireless Sensor Networks
SUN Lijuan, Yang Guannan, GUO Jian, XU Wenjun, WANG Ruchuan
2010, 19(4): 589-593.
Abstract(796) PDF(1637)
Abstract:
Energy of nodes is limited in wireless sensornetworks, so it is necessary for routing protocol to makeuse of energy effectively. Aiming at the disadvantages ofdirected diffusion in energy utilization, a routing protocolbased on ant colony algorithm for wireless sensor networks(Directed diffusion based ant colony algorithm, DDBA) ispresented. This protocol, based on the principle of antcolony algorithm, considers both the residual energy ofnodes and the transmission energy consumption betweennodes when routing. Theoretical analysis and simulationresults show that DDBA reduces the power consumption,enhances the energy utilization and balances the energyconsumption of the whole network, which leads to extensionof the lifetime of both nodes and networks.
TAPEMAN: Towards an Optimal Data GatheringMechanism in Wireless Sensor Networks
GU Yu, JI Yusheng, CHEN Hongyang, ZHAO Baohua
2010, 19(4): 594-598.
Abstract(579) PDF(680)
Abstract:
In this paper, we study the lifetime optimizationproblem using a mobile sink node in a storageconstrainedwireless sensor network, and propose an optimaldata gathering mechanism named (TAPEMAN) whichruns in three steps: first a tight upper bound of the networklifetime is derived through analysis of energy consumption,then we generate a Traveling salesman problem(TSP) solution based on a 2-approximation O(n2) algorithm.A judgment will be made about whether thissolution is optimal. If the answer is yes, TAPEMAN terminates.If not, a novel data diffusion scheme is used todistribute data to neighboring nodes, in order to avoid dataleaking. We prove that under some reasonable assumptions,our algorithm can achieve this upperbound. Simulationresults demonstrate the efficiency of our proposedsolution and substantiate the importance of using sink mobilityfor energy-constrained sensor networks.
CSa-XCHOKe: A Congestion Adaptive CHOKeAlgorithm
JIANG Ming, WU Chunming, ZHANG Min, BIAN Hao
2010, 19(4): 599-602.
Abstract(572) PDF(1028)
Abstract:
To solve the problem of fair bandwidthsharing between responsive flows and unresponsive flows,this paper proposes a new algorithm named CSa-XCHOKe,which identifies unresponsive flows by using CHOKe hitshistory records. In this algorithm, the number of pickedpackets from queue is decided by the congestion level, andthe picked packets are compared with the arrival packetto judge whether they belong to the same flows. If theyhit, the packet drop probability is calculated based on congestionlevel and link load; otherwise, CSa-XCHOKe usesMRED scheme to process the arrival packet. The performanceof CSa-XCHOKe is compared with other mainschemes like XCHOKe, CSFQ and CHOKe, and the simulationresults show that CSa-XCHOKe performs betterthan other schemes in bandwidth sharing fairness issues.
Efficient Workload Adaptive Scheme forLow-power Real-time Systems
ZHANG Zhe, QIAN DejunZHANG Zhe, HU Chen, QIAN Dejun
2010, 19(4): 603-607.
Abstract(582) PDF(669)
Abstract:
Low-power schemes for real-time systemsthat dynamically vary CPU speed and voltage dependingon the probability distribution of the task’s work requirementwere implemented in previous studies to save energy.However, they did not consider the interrelationshipof the probability distribution between running tasks’work requirement to further reduce the power consumption.This paper presents a new low-power DVS schemeto combine all running tasks’ scheduling into a single optimizationproblem such that the frequency schedule of eachtask is determined by the system workload. The computationof the frequency assignment is optimized based on theprobability distribution of the workload’s clock demands.This DVS scheme keeps the real-time system schedulablebecause the longest time duration to execute these clockcycles is restricted to meet all tasks’ deadlines. Simulationresults show that it effectively reduces at least 30% lessenergy than previous studies without any negative effecton performance.
A Markov Decision Model for Low-LayerReadaheads
XU Chenfeng, XI Hongsheng, WU Fengguang
2010, 19(4): 608-612.
Abstract(556) PDF(628)
Abstract:
Readahead is an important technology forimproving IO performances. Its performance rests withthe prefetching policy. We introduced a Markov decisionmodel to describe low-level read processes and readaheadbehaviors such as those in modern operating system kernelswhen there is enough memory for caching. By usingthis model, it is possible to analyze various readaheads andfind better prefetching policies for specific read patternswith fewer practical tests. For illustration, we presentedan example about the ondemand readahead in linux kernel.The experiments show the model agrees with the realreadahead and the found policies significantly outperformthe current in real world.
A Novel Two-Phase XML Keyword QueryAlgorithm
LIN Xudong, WANG Ning, XU De
2010, 19(4): 613-617.
Abstract(635) PDF(1011)
Abstract:
Many existing XML keyword query approachesadopt the subtrees rooted at the smallest lowestcommon ancestor of the keywordmatching nodes as the basicresult units. The structural relationships among XMLnodes are excessively emphasized in these approaches butthe context meanings of XML nodes are not taken seriously.To change this situation and improve the matching betweenusers’ query intentions and final query results, wepropose a two-phase XML keyword query algorithm. Inthe first phase of the algorithm, users can select the suitablecontext meanings of keyword matching nodes to matchtheir query intentions, and the eligible keyword matchingnodes will be found in the second phase to return moreaccurate query results efficiently. The effectiveness andthe efficiency of the algorithm are demonstrated throughextensive experiments.
A Model for Virtual Corporations CooperativePartner Selection Based on Hybrid GeneticAlgorithm
LI Jian, JING Bo, CHEN Xiubo, WANG Cong
2010, 19(4): 618-620.
Abstract(584) PDF(688)
Abstract:
To enhance the efficiency of corporationcooperative partner selection and optimization process invirtual corporations, a Hybrid genetic algorithm (HGA)and its application in cooperative partner selection processand optimizations model for was presented. In themodel, the hybrid genetic algorithm is used to enhancethe efficiency of corporation cooperative partner selectionand optimization process. In the experiments, four kindsof genetic algorithms are used to compared with, they areStandard genetic algorithm (SGA), the genetic algorithmbased on metropolis rule (MGA), Adaptive genetic algorithm(AGA) and HGA. After 1000 times of experimentsto gain the optimal result, SGA averagely needs 168 runsof negotiations, MGA averagely needs 146 runs of negotiations,AGA averagely needs 125 runs of negotiations,while HGA averagely needs only 101 runs of negotiations.The experimental results show that the HGA can gain theoptimal result more efficiently than other four three kindsof genetic algorithm in corporation cooperative partner selectionand optimization process.
A Secure Communication Mechanism of P2PRFID Code Resolution Network
LIU Xueyang, ZHAOWen, HUANG Kaimu, FENG Zhiming, ZHANG Shikun, WANG Lifu
2010, 19(4): 621-626.
Abstract(633) PDF(999)
Abstract:
Due to the characteristics of P2P network,the correctness of P2P RFID code resolution service is notguaranteed. In this paper, a novel secure communicationmechanism for the P2P RFID code resolution network isproposed based on double random numbers. In this mechanism,the nodes involved in the resolution process generaterandom numbers, and the random numbers are used fornode authentication and message authentication. Based onKademlia, the proposed secure communication mechanismin the P2P RFID code resolution network is implemented.In experiments, the P2P RFID code resolution networkwith the proposed secure communication mechanism is notonly load balancing, extensible, and node failure tolerant,but also highly efficient in code resolution. At last, thefuture work is presented.
Efficient Scheme for Path Recognition fromIntegrated Circuit Layout
YUAN Zhaohui, SUN Shilei, WANG Gaofeng
2010, 19(4): 627-630.
Abstract(602) PDF(712)
Abstract:
An efficient scheme for the centerline-basedpath recognition from an IC mask layout is presented. Unlikethe division-based methods, a tree-traverse-based approachis proposed. This new scheme can be realized as areverse procedure of the layout generation from wire routingtrees. Moreover, this scheme can handle complex allanglewires. Experimental results show that this schemehas nearly linear computational complexity yet generatesprecise results.
Structural and Photoelectrical Properties of NiOThin Film
XIANG Rong, WANG Xin, JIANG Delong, LI Junqiao, WANG Guozheng, LI Ye, DUANMU Qingduo, TIAN Jingquan
2010, 19(4): 631-633.
Abstract(536) PDF(1594)
Abstract:
NiO thin film is a p-type wide-bandgapsemiconductor material and has wide applications in magneticdevice, chemical sensor, especially in Ultra-violet(UV) detector devices. In this paper, NiO thin films weredeposited on quartz-glass substrate by magnetron sputteringmethod. The influences of the sputtering voltageon the structural, optical and electrical properties of NiOthin film were mainly investigated. It was found that thestructural, optical and electrical properties of NiO thinfilm were greatly dependent on the sputtering voltage. Atlower sputtering voltage, the NiO thin film was of amorphousstructure with no preferred orientations; the amorphousNiO thin film changed into crystalline one at highersputtering voltage. With the increasing of sputtering voltage,the optical transmittance became poorer, the bandgapand the resistivity of NiO thin film decreased. These obtainedresults will be helpful for optimizing the fabricationprocess of NiO thin films, ultimately to obtain high qualityNiO-based devices.
A Memory-efficient Multi-dimensionalHardware-specific Algorithm for PacketClassification
HUO Hongwei, YE Mangu, GAO Dongpei
2010, 19(4): 634-636.
Abstract(731) PDF(761)
Abstract:
The paper presents a memory-efficientmulti-dimensional hardware-specific algorithm for fastpacket classification. The algorithm builds a decision treein which each leaf node stores a relatively small numberof rules. The maximum number of rules is determined bythe level of a node in the tree and the maximum availablesearching time so that the worst-case classification timecan be bounded. The algorithm allows quick updates andhas relatively small storage requirements. It can be tailoredfor a Field-programmable gate array (FPGA) implementationusing an optimization for the tree and a simplememory management strategy. The results show that thealgorithm can classify about 2.5M packet headers per secondon 50MHz search clock with the worst-case classificationtime Csum = 34 clock cycle and the space complexityO(n).
The Energy Optimization for Architectures withLimited Addressing Modes Using Scratch-PadMemory
LING Ming, ZHANG Yang, MEI Chen, PU Hanlai
2010, 19(4): 637-640.
Abstract(615) PDF(690)
Abstract:
Low power design plays a very importantrole in the modern embedded system. Compared withtraditional cache, many researchers focus on the substitute:Scratch-pad memory, which can reduce energy consumptionand guarantee the overall performance. However,few of these optimization schemes took architectureswith limited addressing modes into consideration, whichmeans ignoring the long-branch overheads existing in almostall of Reduced instruction set computer (RISC) architectures.This paper implements two matrices to illustratethose overheads quantitatively, and eventually figuresout, in virtue of our improved knapsack algorithm and itscorresponding dynamic programming, the most optimizedimage layout for energy considerations. Compared withgeneral SPM optimization method, experiments achieve upto 58.6% and average 19.5% decrease in energy consumptionwithout any performance degradation when SPM sizeis only 8kbytes.
Multi-Class Maximum A Posteriori LinearRegression for Speaker Verification
ZHANG Xiang, XIAO Xiang, WANG Haipeng, ZHANG Jianping, YAN Yonghong
2010, 19(4): 641-645.
Abstract(569) PDF(909)
Abstract:
Maximum likelihood linear regression(MLLR) transforms have proven useful for textindependentspeaker recognition systems. These systemsuse the parameters of MLLR transforms as features forSVM modeling and classification. In this paper, we focuson calculating affine transforms based on a GMMUniversalbackground model (UBM). Rather than estimating transformsusing maximum likelihood criterion, we propose touse Maximum a posteriori linear regression (MAPLR) forfeature extraction. This work is enriched by a multi-classtechnique, which clusters the Gaussian mixtures into regressionclasses and estimates a different transform foreach class. The transforms of all classes are concatenatedinto a supervector for SVM classification. Besides, a furtheraccuracy boost is obtained by combining supervectorsderived from both female and male UBMs into a largersupervector. Experiments on a NIST 2008 SRE corpusshow that the MAPLR system outperforms MLLR andthe multi-class approaches can also bring significant gains.
Image Denoising Using Bandelets and HiddenMarkov Tree Models
ZHANG Wenge, WANG Suang, LIU Fang, GAO Xinbo, JIAO Licheng
2010, 19(4): 646-650.
Abstract(649) PDF(831)
Abstract:
In this paper, both the marginal and jointstatistics of second generation Orthogonal bandelet transform(OBT) coefficients of natural images are firstly studied,and the highly non-Gaussian marginal statistics andstrong interscale, interlocation and interdirection dependenciesamong OBT coefficients are found. Then a HiddenMarkov tree (HMT) model in OBT domain which can effectivelycapture all dependencies across scales, locationsand directions is developed. The main contribution of thispaper is that it exploits the edge direction information ofOBT coefficients, and proposes an image denoising algorithm(B-HMT) based on HMT model in OBT domain.We apply B-HMT to denoise natural images which contaminatedby additive Gaussian white noise, and experimentalresults show that B-HMT outperforms the Wavelet HMT(W-HMT) and Contourlet HMT (C-HMT) in terms of visualeffect and objective evaluation criteria.
A Mesh Partitioning Approach for 3D MeshOblivious Watermarking
YAO Zhiqiang, PAN Rijing, LI Fenghua, WENG Bin
2010, 19(4): 651-655.
Abstract(611) PDF(672)
Abstract:
Mesh partitioning approach, widely used in3-D meshes compression and coding, is able to be extendedto exploit local features and embed watermark string repeatedly,and thus it may improve watermarking robustness.This paper presents a segmentation method whichcan predict the partition boundary of stego-mesh with asmall amount of additional data. The face normal principalcomponent analysis is employed to determine the meshes’canonical positions. Some additional scattered informationfrom the elevation of cover-mesh in canonical coordinatesystems, instead of the presence of the cover-mesh, ispicked up to match the elevation of stego-mesh. With theaim to realign the stego-mesh approximately and plot outthe partition boundary, the deflected degree of coordinatesystem is estimated. Experimental results show that thenew partitioning scheme is stable against the cropping andsimplification attacks.
Harmonic and Interharmonic Signal AnalysisBased on Generalized S-Transform
QUAN Huiming, DAI Yuxing
2010, 19(4): 656-660.
Abstract(552) PDF(755)
Abstract:
This paper presents a new approach to analyzeharmonics using the Generalized S-transform (GST)and clonal selection algorithm. After Comparing the effectof p and k in GST on the harmonics detection, a modifiedGST was adopted, in which the k parameter waschanged with the different frequency band, and the frequenciesof interharmonics detected mistakenly were correctedthrough the instantaneous frequencies. The harmonicsamplitudes and phases given by GST were withsome error, so the clonal selection algorithm was used tooptimize them, and the detail optimization process wasdescribed in this paper. The simulation results show thatthe method could analyze the harmonics and interharmonicsaccurately, and its performance is far better than FFTespecially in analyzing the interharmonics and transientharmonics.
Linguistic Steganalysis Based on Meta Featuresand Immune Mechanism
YANG Hao, CAO Xianbin
2010, 19(4): 661-666.
Abstract(553) PDF(1391)
Abstract:
Linguistic steganalysis depends on efficientdetection features due to the diversity of syntax andthe polysemia of semantics in natural language processing.This paper presents a novel linguistics steganalysisapproach based on meta features and immune clone mechanism.Firstly, meta features are used to represent texts.Then immune clone mechanism is exploited to select appropriatefeatures so as to constitute effective detectors.Our approach employed meta features as detection features,which is an opposite view from the previous literatures.Moreover, the immune training process consistsof two phases which can identify respectively two kindsof stego texts. The constituted detectors have the capableof blind steganalysis to a certain extent. Experimentsshow that the proposed approach gets better performancethan typical existing methods, especially in detecting shorttexts. When sizes of texts are confined to 3kB, detectionaccuracies have exceeded 95%.
A Virtual Entertainment Robot Based onHarmonic of Emotion and Intelligence
HUANG Xiangyang
2010, 19(4): 667-670.
Abstract(547) PDF(800)
Abstract:
A virtual robot called Hubot (Humanoidrobot) is proposed to enhance entertainment and to highlypersonalize human-machine interaction. The model ofHubot integrates emotional system with intention system.The extended situation calculus including affective computingis used to describe the model. Fuzzy representationis used to capture the inherent uncertainty of emotionand cognition. Finally, the correctness and validity of themodel is proven by an experiment of comparison.
A Hessian-Laplace Based Blob Detector forStamp Image Classification
MING Anlong, MA Huadong
2010, 19(4): 671-675.
Abstract(612) PDF(968)
Abstract:
Blob features are usually used as texturedescriptors for object classification and most traditionalblob detectors nowadays are luminance-based. However,in many applications we could require a blob detector forthe color domain. In this paper, we propose a novel andefficient framework for the extension from scalar-signals tovector-signals of multi-scale blob detection to prevent informationloss due to gray scale transformation. Then, wepresent a blob detector for stamp image classification todistinguish the stamp types. In our experiments, we compareour method to other approaches. The experimentalresults demonstrate the effectiveness of our proposed detector.
Example-based Learning for Depth Estimation ofFacial Landmarks and Its Application in FaceModeling
GONG Xun, WANG Guoyin
2010, 19(4): 676-680.
Abstract(563) PDF(836)
Abstract:
When reconstructing a 3D face from a singleimage, the unknown depth of landmarks degrades themodel’s accuracy considerably. A promising solution is topredict the depth of landmarks by learning from 3D examplesof scan before modeling. This paper proposed to usea sparse linear model to estimate the depth of landmarksfrom their prior distributions in a 3D face database. Theestimated 3D landmarks were applied to the deformationprocess to ensure a more precise facial shape for a given image.Tests on synthesized images show that the estimatedfeatures’ depth is closely to the known ground truth andthat the modeling accuracy of various deforming methodsis greatly enhanced with the estimated 3D features.
Low Complexity Iterative Joint Source-ChannelDecoding of Variable-Length Coded CorrelatedSources
YANG Yang, YI Weidong
2010, 19(4): 681-686.
Abstract(535) PDF(654)
Abstract:
The paper proposes a Soft-input-softoutput(SISO) A posteriori probability (APP) decoding algorithmfor variable-length coded correlated sources. Thistechnique can obtain Maximum likelihood (ML) sequenceestimation and bit-based reliability. The notable featureof this technique is its low computational complexity. Anotherinnovation of the paper is that, based on the Extrinsicinformation transfer (EXIT) charts analysis, we obtainedconstant Scaling factors (SFs) to refine the extrinsicinformation in the Iterative joint source-channel decoding(ISCD) process. This technique improves the ISCD performancewith adding very limited computational complexity.Simulation results showed that the proposed ISCD schemeprovided a great improvement on error protection capabilityfor the variable-length coded correlated sources.
An Improved Geometric Active Contour Modelfor Concrete CT Image Segmentation Based onEdge Flow
HE Ning, LU Ke, BAO Hong
2010, 19(4): 687-690.
Abstract(577) PDF(780)
Abstract:
To put forward an improved Geometricactive contour (GAC) model based on the edge flow forsegmentation of concrete Computed tomography (CT) images.An edge flow vector field is constructed at each pixellocation basing on the integration of image intensity, color,and texture characteristics to make the vector field point tothe direction of potential boundary pixels. The GAC fieldobtained by the diffusion of the edge flow vector is regardedas the external potential force of the GAC model. Then thecurve propagation is guided by the improved GAC model.The experimental results obtained by applying the methodto segment concrete CT images. By the integration of theedge flow into the GAC model, the segmentation qualityis obviously improved.
An Efficient and Scalable Locator/ID SeparationScheme Based Mobility and MultihomingSupporting Architecture for 3G SAE
YAN Zhiwei, ZHANG Hongke, ZHOU Huachun, ZHANG Sidong
2010, 19(4): 691-698.
Abstract(524) PDF(896)
Abstract:
System architecture evolution (SAE) is oneof the key challenges for the Long term evolution (LTE) of3G research, aiming to develop a new core network forimproving the IP based packet-switched network performance.While for the current IP-based mobility managementarchitecture in 3GPP SAE, hosts are identified byIP addresses that depend on their topological location. Inother words, the IP addresses are semantically overloadedsince they identify both hosts and topological locations.In this paper, on the basis of an efficient and scalableLocator/ID separation (LIDS) scheme, a mobility and multihomingsupporting architecture for 3G SAE is designed.From the qualitative and quantitative analyses of the 3GPPSAE recommended mobility management schemes (i.e.,Mobile IPv6 and Proxy Mobile IPv6) and our architecture.We show that our LIDS based architecture providesan outstanding network-based mobility management withroute optimization support and lower signaling cost thanIP-based schemes. Furthermore, a comprehensive comparisonamong the various well-known Locator/ID separationbasedarchitectures is made.
A Dynamic Programming Based ReplicaPlacement Algorithm in Data Grid
FU Xiong, WANG Ruchuan, DENG Song
2010, 19(4): 699-704.
Abstract(549) PDF(675)
Abstract:
Data replication is one of the well-knowntechniques to reduce the cost of data access and networkbandwidth consumption, as well as to improve data availabilityin data grid. The challenge in replica managementis to select a set of suitable nodes for replicas and it isknown as replica placement problem. In this paper, we addressreplica placement problem in data grid under giventraffic pattern and propose a replica placement algorithmbased on dynamic programming. It can help us find an optimalnode set for replicas subject to workload constraintof all replicas while total communications cost in data gridis minimized. Numerical example and experiments showthat our algorithm is feasible and effective for the replicaplacement problem.
Unlinkable Secret Handshakes from MessageRecovery Signature
WEN Yamin, ZHANG Fangguo, XU Lingling
2010, 19(4): 705-709.
Abstract(589) PDF(773)
Abstract:
Secret handshake scheme allows the membersof a certain organization can anonymously authenticateeach other. In this paper, two Unlinkable secret handshakeschemes (called USH-1 and USH-2) are proposedby using the Message recovery signature (MRS). USH-1 achieves the unlinkability with one-time pseudonyms,whilst USH-2 obtains a strong unlinkability against groupauthority with reusable credentials. The security of USH-1 and USH-2 are reduced to the intractability of DiscreteLogarithm Problem and the k + 1 Square Roots Problem,respectively. Compared with some seminal schemes, bothof our schemes are competitive in the performance.
Self-generated Anonymous Credentials in Ad HocNetworks
ZHANG Yang, CHEN Junliang
2010, 19(4): 710-716.
Abstract(572) PDF(848)
Abstract:
Classical anonymous credential systems aredesigned for center-controlled networks, and not applicablefor ad hoc networks. The existing pseudonym schemesdo not thoroughly address the issue in ad hoc networks.Therefore, this paper proposes a self-generated anonymouscredential system for ad hoc networks, where a node generatesvalid pseudonyms itself without the participation ofan organization management center, and the frequency ofpseudonym-generation is controllable. The formal specificationof the proposed system is not defined based onzero-knowledge protocols and classic pseudonym models,but based on the ideal functionality for common certification.To implement a practical system, a modular methodologyis used, where a pseudonym-based signature schemeis proposed, and then a tag-based signature protocol ispresented. Finally, the whole system is constructed. Thesecurity of our scheme is proved in the ideal-system/realsystemmodel.
Optimal Amplify-and-Forward Relaying Schemesfor Two-hop MIMO Relay Networks
FU Youhua, YANG Luxi, ZHU Weiping
2010, 19(4): 717-722.
Abstract(587) PDF(804)
Abstract:
This paper proposes two relaying schemesfor the two-hop and half-duplex Amplify-and-forward (AF)distributed relaying system with one source, one destination,and multiple relays, each equipped with multiple antennas.The maximum transmission rate solution subjectto individual relay power constraint is investigated underan assumption of known Channel state information (CSI).Numerical simulations verifying the analytical results showthat an appropriate relaying processing can largely improvethe system spectral efficiency.
Weakness of Kurosawa and Heng’s WitnessIndistinguishable Protocol Based on DH-tuple
ZHOU Yajian, ZHANG Juanmei, LI Zichen
2010, 19(4): 723-726.
Abstract(535) PDF(763)
Abstract:
The concept of witness indistinguishabilityand witness hiding was introduced by Feige and Shamirin 1990. Recently, B. Kurosawa and S.H. Heng proposed3-move undeniable signature scheme at Eurocrypt 2005,which has the confirmation and disvowal protocols forproving the message-signature pair is or is not valid respectively.They also proposed an new DH-tuple Witnessindistinguishable (WI) protocol, which is the foundationof the comfirmation protocol in this new undeniable signaturescheme. In this paper, we will first show a weaknessin Kurosawa and Heng’s witness indistinguishable protocol.In general, there are two or more witnesses in a WTprotocol. The weakness in Kurosawa and Heng’s WI protocolis that a prover with a witness, but does not konwanother witness, can forge to produce a cheating proof,which is considered to be from another witness. So, fromthe concept of witness hiding, there will be an new witness,which is difference from the two original witnesses. We alsoinvestigate the reason cuased the weakness in Kurosawa etal.’s protocol.
Three-Stage Based Rapid Long PN-codeAcquisition Method by EmployingTime-Frequency Folding Technique
LI Hong, LU Mingquan, FENG Zhenming
2010, 19(4): 727-732.
Abstract(562) PDF(963)
Abstract:
Rapid long Pseudo-noise (PN) code acquisitionis challenging in Direct sequence spectrum system(DSSS) and Global navigation satellite system (GNSS).The difficulty is how to quickly detect numerous codephases and lots of frequency cells. In the study, a threestageacquisition method is proposed. In search stage one,a few code phases and several frequency cells are mergedinto a single cell, to directly reduce the cardinality of searchspace. The reduction is favorable to accelerate search process.Then, in search stage two and search stage three,fine acquisitions for frequency cells and code phases areperformed, respectively. Compared with the Extendedreplica folding acquisition search technique (XFAST), theproposed method tests more code phases and frequencycells simultaneously and consequently reduces mean acquisitiontime at the expense of degrading detection performance.Numerical results demonstrate the analysis.
Load-Balanced and Length-Minimized LinkScheduling for Multi-Channel TDMA WirelessMesh Networks
JIN Junfeng, ZHAO Baohua, ZHOU Hao
2010, 19(4): 733-736.
Abstract(585) PDF(590)
Abstract:
To minimize the length of scheduling andguarantee the load balance of channels, a Load-balancedand length-minimized link scheduling (LBLM) algorithmis proposed. LBLM algorithm is a heuristic scheme, whichassigns time slots for unicast traffic based on link’s weightand hop-count in the routing traffic tree. Thus the algorithmconsiders both primary and secondary interference,as well as guarantees the proportional fairness. Thens2 simulation results show that in multi-channel TDMAWireless mesh networks (WMNs), the proposed algorithmhas the benefits of lower complexity, shorter frame lengthand better channel balance compared to other well-knownschedule mechanisms.
Joint Estimation of Synchronization Parametersand Channels for Generalized-OFDMA Uplink
GUO Junqi, SHANG Yong, REN Shubo, XIANG Haige
2010, 19(4): 737-742.
Abstract(588) PDF(842)
Abstract:
We propose an estimation algorithm tojointly estimate frequency/timing synchronization parametersand channels for the uplink transmission of Orthogonalfrequency division multiple access (OFDMA) companiedwith a generalized carrier assignment scheme (generalized-OFDMA). First, the proposed algorithm employs a Zadoff-Chu (ZC) sequence which exhibits an ideal circular autocorrelationproperty. The different circular-shifting formsof this ZC sequence are allocated to different users as theirtraining sequences, respectively. Furthermore, the proposedalgorithm implements correlation operations on themixed training signal received by the base station so as tojointly estimate Carrier frequency offsets (CFOs), Timingerrors (TEs) and Channel impulse responses (CIRs) of allusers. Simulation results show that the proposed algorithmnot only achieves joint CFO/TE/CIR estimation, but alsoprovides better CFO, TE and CIR estimation performancethan some existing estimation algorithms for OFDMA uplink.
A Multiscale Maximum Entropy ReconstructionAlgorithm for Synthetic Aperture ImagingRadiometers
ZHOU Xi, HU Weidong, SUN Houjun, LI Bo, LV Xin
2010, 19(4): 743-746.
Abstract(609) PDF(1316)
Abstract:
This paper introduces a multiscale maximumentropy image reconstruction algorithm for twodimensionalsynthetic aperture imaging radiometers. Thenovelty of the method is to use wavelet decomposition tocontrol the noise amplification in the different scales. Andour work is focused on the application and the adjustmentof the Maximum entropy method (MEM) to the Earthobservation situation. Several numerical simulations havebeen performed for synthetic aperture imaging radiometersin the Geosynchronous Earth orbit (GEO). The simulationresults indicate that, with the proposed method, itis feasible to improve the spatial resolution of the reconstructedimage while preserving the radiometric sensitivitysimultaneously.
About the Ambiguities in DOA Estimation
LIU Tingting, XIAO Ling
2010, 19(4): 747-751.
Abstract(544) PDF(2057)
Abstract:
Ambiguous set calculation has been paidmuch attention in signal processing related areas. Existingconstructing methods can only get one type of ambiguoussets. In this paper, an iterative numerical method isproposed to obtain the nearby ambiguous sets from an arbitraryinitial value. Furthermore, for linear array withsome symmetric sensors, a combination method based onthe non-uniform partition of the array manifold is also suggestedto get ambiguous sets. Representative examples arepresented to show the effectiveness of the proposed methods.In addition, the case of non-uniform basic set hasbeen found for non-symmetric array from the examples.
A Novel Maneuvering Target Tracking Algorithmfor Radar/Infrared Sensors?
YIN Jihao, CUI Bingzhe, WANG Yifei
2010, 19(4): 752-756.
Abstract(677) PDF(720)
Abstract:
In this paper a novel filtering algorithm,called Radar/Infrared(R/IR) converted measurementsfilter Interacting multiple model (R/IRCMF-IMM),is proposed for tracking a maneuvering target usingRadar/Infrared heterogeneous sensors. This filtering algorithmis developed by converting the polar measurementsof Radar and Infrared to Cartesian coordinates, and calculatingthe statistic characteristics of converted measurementerrors before filtering, then applying to the IMMtechnique. Since there are no linearization errors of themeasurement model in the process, the new method hasbetter tracking performance than traditional IMM that usingExtend Kalman filters (EKFIMM). Additionally, thenew method has equal calculation cost with EKFIMM. Finallya simulation example is given and shown that the proposedalgorithm achieves significant improvement in theaccuracy of tracking estimation comparing with the EKFIMM.
Ion Feedback Suppression for Microchannel PlateApplied to Third Generation Image Intensifiers
PAN Jingsheng, LV Jingwen, ZHANG Zhengjun, SUN Jianning, SU Detan
2010, 19(4): 757-762.
Abstract(632) PDF(2039)
Abstract:
The application of an Ion barrier film (IBF)over the input side of the Microchannel plate (MCP), toprotect the Gallium Arsenide (GaAs) photocathode fromdamage by the ion feedback from MCP during operation,while provide third generation (Gen. 3) image intensifierwith longevity of service, also cause which sufferd a significantdegradation in Signal to noise ratio (SNR). Thus,for improventment of the SNR of Gen. 3 image intensifier,the essential way is to suppress the generation of ionfeedback within MCP in order to thin the IBF or evencompletely removal of it. In this paper a MCP based ona glass-ceramic substrate was developed, this MCP withoptimizations of glass composition and devitrification remodificationof microstructure which show great promiseto suppress the generation of ion feedback within MCP,making it possible to thin the IBF and fabricate a Gen.3 image intensifier with improved SNR while maintainingreliability requirement.
On the Use of Multi-Tone for the Estimation andMeasurement of Noise Power Ratio inThird-Order Nonlinear System
HUANG Cheng, WU Jianhui, SHI Longxing
2010, 19(4): 763-768.
Abstract(589) PDF(990)
Abstract:
Multi-tone signal is widely used to estimateand measure the nonlinear distortion in communicationsystem. A generalized formula for the estimation ofNoise power ratio (NPR) distortion in memoryless thirdordernonlinear system is presented in this paper. An accurateand low-cost NPR measurement setup with fewertones than traditionally used is also introduced in this paper.Accurate measurement of NPR distortion is achievedby averaging distortion power stimulated by multi-tone signalswith different random phases. Automatic measurementprocedure is developed to reduce the test time andapply the test setup for the production environment. Comparingthe measurement results between our method using60-tone and traditional technology using 10,000-tone,the NPR measurement error is only 0.23dB. The accuracyof the NPR estimation formula is verified using this testenvironment. The variance between estimation and measurementis below 1dB when the DUT is in the lightlynonlinear region.
Micro-Doppler Signal Analysis Based on theGeneralized S Transform for Radar Targets
WANG Jun, LEI Peng, SUN Jinping, MAO Shiyi
2010, 19(4): 769-774.
Abstract(625) PDF(1269)
Abstract:
The Generalized S transform (GST), whichcombines advantages of the Short time Fourier transform(STFT) and the Continuous wavelet transform (CWT), isan effective method to analyze and extract time-frequencyfeatures. On the basis of micro-Doppler effect in radarecho, an improved GST with adjusted window functionwas introduced, and its expression as operation of theFourier spectrum was developed. Simulated radar signalsof typical micro-Doppler modulation were applied tocompare and test the GST and other three common timefrequencytechniques. The experimental results show thatthe GST with adjusted window function is a noise robustand adaptive transform, reduces computational burden,and provides more details of the local time-frequency featuresused to improve resolution and parameter estimationof micro-motion, which demonstrates its effectiveness inmicro-Doppler signal analysis.