摘要:The edge details in images involve significant visual perception information,which play an important role in the further image understanding and scene perception.An image edge detection and extraction method is presented based on the fractional differentiation theory aiming at solving the problems of inaccurate edge detection like fractal structures in the images by the traditional edge detection methods.Firstly,the inflexion points in the images are detected based on the characteristics of fractional differentiation;then,an image edge detail detection and extraction operator with sub-pixel interpolation is derived from the Grumwald-Letnikov (G-L) definition of fractional differentiation combining with the Lagrange interpolation polynomials.The experimental results demonstrate that the proposed operator is capable of extracting image edge details (inflexion points) efficiently.Furthermore,it is able to detect the useful object edge details from image with serious noise to achieve better visual effect.
摘要:For the MIMO systems with a large number of antennas and a large QAM constellation,the existing parallel detection algorithms,which are based on channel partition,can approach the optimal detection performance but the computational efficiency is sacrificed.In order to solve the problem,a new parallel detection algorithm based on channel partition is proposed.This algorithm can not only efficiently reduce the computational complexity but also guarantee the detection performance.After channel partition,this algorithm firstly employs the lattice reduction algorithm under the MMSE criterion to improve the properties of the sub-channels;and then the remaining signals are detected with a nonlinear method.The simulation results show that the proposed algorithm achieves the near-optimal detection performance for the 4*4 and 6*6 MIMO systems and its performance decreases 1dB for the 8*8 MIMO systems compared to the optimal performance.Complexity analysis shows that its complexity is reduced 90% or above for 6*6 MIMO systems and 98% or above for 8*8 MIMO systems at the same BER.
关键词:multiple-input multiple-output(MIMO);minimum mean square error(MMSE);channel partition;parallel detection;lattice reduction;QAM
摘要:In order to improve the accuracy and efficiency of the topology inference algorithm for unicast network,an efficient and adaptive topology inference algorithm is proposed.With the information of TTL hop count,this algorithm reduces the number of the probe pairs needed in the process of bisection Depth-First Search Ordering,and improve the efficiency of the topology inference.On the other hand,through the analysis of the principle of the Depth-First Search topology inference algorithm,a sufficient condition for the algorithm to return the correct network topology is given.Based on this condition,an adapt threshold selection method is proposed,it can improve the accuracy of the topology inference when the network link parameters are unknown.Simulation results show this algorithm can obtain a higher accuracy and efficiency.
摘要:A new method for the restoration of degraded images in inhomogeneous atmosphere was proposed using the optical model of image degradation in inhomogeneous path.Firstly,Sky radiances at observer corresponding to the slant directions of each object in image were estimated.Then,based on the prior knowledge of dark channel,the rough relative distances of scene objects in image were estimated approximately,and as a key point,the ratio factor which reflects the inhomogeneous property of visual path in the degradation model was obtained,and subsequently atmospheric transmittance of imaging path was computed.Finally,using these parameters computed above,the degraded image can be restored by the inverse process of the optical degradation model.The experimental results show that the new algorithm is effective for actual outdoor degraded images.
关键词:image dehazing;inhomogeneous atmosphere;image degradation;optical model
摘要:In the process of color image edge detection,it is very difficult for the extraction of detail edges in the areas with low illumination.A novel method of color image edge detection based on modified visual perception model is proposed in this paper.Because hue component is represented by circular data,it easily leads to the false edges.Firstly,we use hue distance to calculate the gradient of hue component,and then use non-linear function whose variant is saturation to control the change of gradient of hue component.Finally,the edges of color image can be detected by computing vector gradient in modified visual perceptual model.Experimental results show that the proposed method can obtain higher accuracy of edge detection than the existing approaches.Especially,the proposed method has a good response for edge details in areas with low illumination.
摘要:An image denoising algorithm is proposed with an overcomplete set of linear transforms for removing white Gaussian noises of the image.The denoising result is gotten by weight averaging.The method computes the weight with the sparse concentrate index.The weight depends on sparse decomposition of localized transform coefficients.Sparser the image region is,the larger its weight is.The method has no need to devise elaborate statistic models on the transform coefficients and more sophisticated transforms for image singularities.The method is simple and obtains very high denoising performance,especially for the image with singularities.Experiments results show the effectivity of the method.
摘要:A novel affine invariant curve descriptor based on membrane vibration model is proposed in this paper.This paper focused on two kinds of curves,of which one is Jardon curve and the other is open curve.For the Jardon curve,it was firstly normalized to overcome the influence of affine distortion.The curve descriptors were then obtained by calculating the eigenvalues of Helmholtz equation using the finite element method;For the open curve,its narrowband model region was firstly established by use of Euclidean distance transform.Then the open curves were replaced by the narrowband models.Finally the open curve descriptors were obtained by using the Jardon curve's method.The experimental results show that the proposed method outperforms the existing Fourier descriptor and Zernike moment descriptor.
关键词:curve matching;membrane vibration;Jardon curves;Helmholtz equation;narrowband model
摘要:In this approach,three conjugate point rules between two overlapped surfaces are discussed,based on a general least squares 3D surface matching model.We explain how these methods can be used for the co-registration of two real 3D point sets,and show co-registration results based on airborne laser scanner data.Concluding results of our experiment suggest that every proposed conjugate point rules has a good performance for transformation parameters estimation in 3D surface matching procedure.In many cases,the least normal distance can achieve best result in practical matching.
摘要:Multiple Input Multiple Output-Orthogonal Frequency Division Multiplexing (MIMO-OFDM) is one of the key technologies in the next generation broadband wireless communication systems.It is very sensitive to timing synchronization,which includes frame synchronization and symbol synchronization.Distributed antenna system (DAS) is considered as an open architecture for future communication systems.With the DAS architecture,the frame synchronization becomes more difficult and more complex.Therefore,it is in urgent need of a performance analysis method to guide the frame synchronization algorithm design.This paper proposed an accurate performance analysis method for frame synchronization in distributed MIMO-OFDM system.The proposed method,which is a Gaussian performance analysis method based on the statistical properties of frame synchronization metric function,is introduced by orthogonally decomposing some signals into the direction angle of dominated signal.The correct frame synchronization probability is calculated by using the 3 sigma principle of the normal distribution.Finally,the accuracy and effectiveness of the proposed method is examined by its applications to different transmitting antenna numbers and several existing algorithms.
摘要:In this thesis,we proposed a method to construct the feedback functions of the nonlinear shift registers based on three items' primary polynomial.First of all,we proved the characteristic state sets of the linear feedback shift registers with the feedback functions of three items' primary polynomial,we extracted its eigenfunction and synthesized the feedback function for the nonlinear shift register.Secondly,we implemented the non-linear feedback shift register on FPGA.Finally,we conducted analyses and calculations for new sequences generated by the nonlinear shift registers and simulated the sequences in direct sequence spread spectrum code division multiple access communication system(DS-CDMA).The results show unanimously that the new sequences dose not only possess ideal pseudorandom property and better linear complexity,but also get lower bit error rate(BER) than m sequence or gold sequence in the same communication system.
摘要:To evaluate the security of a class of generalized Feistel ciphers,the security evaluation against differential and linear cryptanalyses is investigated deeply using iterative structure.If round functions are all bijective,then the number of active round functions for 4r-round (r≥1) ciphers are not less than (8/3)r-[(rmod3)/3]+(rmod3)/3.The result is at least improved 20% than the existing result when r≥6.So the upper bounds of maximum differential characteristic and linear approximation probabilities for 4r-round (r≥1) ciphers can be estimated if maximum differential and linear approximation probabilities for round function are given.
摘要:Fuzzy identity-based encryption can provide an error-tolerance property for identity-based systems,and it allows a sender to encrypt a message to all users who have a certain set of attributes.In this paper,we proposed a fuzzy identity-based encryption scheme with constant size of ciphertexts that is indistinguishable against adaptive chosen ciphertexts attacks(CCA2)in the standard model.Compared with other existing schemes,this construction can satisfy the application requirements,as it can provide better efficiency in terms of the communication and computation cost as well as a stronger security guarantee.
关键词:fuzzy identity-based encryption;chosen ciphertexts attack;bilinear pairing;standard model
摘要:A PN code optimal design based on unequal-length code modulation is proposed in this paper for better performance of code acquiring for weak signals.The principle is that by using un-equal-length code in spread-spectrum modulation the cross-correlations between different signals have been sliding,thus suppressing multiple access interference effectively.Theoretical analysis and simulation show that this design has predictive power for compressing multi access interference.
摘要:This paper proposes a new algorithm based on parallel factor(PARAFAC) analysis for direction of arrival(DOA) and range estimation of near-field cyclostationary sources.We compute three third-order cyclic moment matrices using the uniform linear array outputs.A parallel factor model is constructed in the cyclic statistics domain and solved via trilinear alternating least squares(TALS).Both DOA and range of each source can be obtained without spectral searching and pairing paramenters.This algorithm can effectively suppress the additive stationary noise with any distribution and the cyclostationary interference.The simulation results of the root mean square error (RMSE) confirm the satisfactory performance of the proposed method.
关键词:wireless location;parameter estimation;direction of arrival(DOA);near-field source;parallel factor(PARAFAC)
摘要:A new DOA(direction of arrival)estimator based on MIMO sonar system is proposed to estimate the DOA of targets.First,orthogonal signals are sent out by transmitting sensors and the output of receiving sensors is matched filter by transmitting signal.Then,solution of space fitting is achieved.Finally,Markov Monte Carlo methods Gibbs sampling is combined with subspace fitting method (GSF) to lighten computation burden.The simulations show that on the condition of two sources,GSF provides similar performance to that achieved by the ML method and better performance than other method in low and mid SNR range but its computational cost is only 1/4 of ML.Moreover,our proposed method is verified by the pool experiment.
摘要:For the purpose of phase center estimation by taking into account the deformation of wing conformal antenna array,a new approach which performed in three stages is proposed.The first stage is carried out at range-Doppler images to estimate the primary eigen-vector associated with a given range-Doppler bin by adopting subspace tracking algorithm.Secondly,the initial estimation of phase center is obtained by fitting the unwrap phase of primary eigen-vector with least square algorithm.Finally,we calculate the deformation parameters based on these initial estimation with total least square algorithm,and the fine estimation of phase center is achieved.The most advantage of the proposed methodology is that none of the auxiliary calibration sources are needed.The effectiveness is confirmed by performance analysis and extensive simulation.
关键词:conformal antenna array;spatial spectrum estimation;phase center estimation;subspace tracking
摘要:This paper establishes the received signal model of parallel bistatic spotlight SAR adopting frequency modulated continuous wave(FMCW),and analyzes the affect of doppler frequency shift introduced by platform's continuous motion while radar transmitting and receiving signals.Then using the concept of instantaneous frequency and through introducing the parameters,i.e.the sum and difference of the closest slant range,and the half quasi bistatic angle,the received signal expression in two dimensional wave-number domain is deduced.Based on this,a two-step processing approach for parallel bistatic spotlight FMCW SAR focusing is presented,which uses azimuth pre-filtering step to overcome azimuth spectral folding phenomenon first.Following this operation,a second residual focusing step is applied to carry out doppler frequency shift compensation and the bistatic frequency scaling processing.Meanwhile,the time frequency relations changing from each stage of azimuth pre-filtering are analyzed in detail.Simulation data processing verifies the analysis and the proposed algorithm.
摘要:In view of the limit of energy in capsule endoscopy,this paper reports a miniature image capsule endoscopy with low power consuming based on wireless power.A wireless energy circuit,an image sensor circuit and a transmitting circuit are designed and studied;a wireless energy transmission system platform is built;a wireless powered image capsule endoscopy prototype is made,and tests are made to prove the rationality of the wireless energy transmission coil.The feasibility of the whole system is proved by image acquisition experiment in the pig intestinethrough cooperation with the hospital.The total size of the image capsule endoscopy prototype is only Φ10mm×20mm,and the power consumption of image sensor circuit and transmitting circuit is 65mW when the modules are connected.
摘要:The interior part of the classic 4-intersection matrix is replaced by the exterior part to represent the RCC5 relations.The 4×4 intersection matrix is then defined to describe the spatial topological relations between a double nesting region and a simple region.It is shown that there are totally 53 topological relations between a double nesting region and a simple region based on three constraints,the illustrations and the conceptual neighborhood graph of all these relations are also given.The reasoning mechanism of this model is established and the composition table is given.The proposed topological relation model can be applied to qualitatively simulate the spatial relations between the typhoon raining areas and the target area,which is useful to some degree on the forecasting and warning of typhoon raining.
摘要:To deal with the limitation of real-time scheduling algorithm of space based optical sensor network,the object characteristics was deeply analyzed,and a sensor pre-assignment model was proposed.Furthermore,the real-time modification was introduced,and the dynamic sensor pre-assignment algorithms based on modified particle swarm optimization and segmented pre-assignment were proposed,under the condition of different demand couple,i.e.tight couple and loose couple.Simulations conducted under the condition of tight couple and loose couple show that,the dynamic sensor pre-assignment algorithm needs long-time pre-assignment process,but the computation efficiency of real-time modification exceeds that of real-time scheduling algorithm,which can leave more margin time for the upper system.
摘要:The problem of localizing multiple odor sources is focused on,and a method of localizing odor sources using robots based on multi-swarm particle swarm optimization(PSO)is presented in this paper.In this method,each robot is regarded as a particle,neighboring particles form a sub-swarm,and different sub-swarms are used to localize different odor sources.In order to make the whole swarm localize as many odor sources as possible,the merging strategy of similar sub-swarms and the reducing strategy of the individual fitness are incorporated into the proposed algorithm.When the environment in which these odor sources lie changes,the globally optimal solution of a sub-swarm is selected according to the relationship between the optimal solution of the sub-swarm in the current generation and that up to the previous generation.The proposed method is applied to localize odor sources in three typical static environments and one dynamic environment,and compared with five previous methods.The experimental results confirm that the proposed method can efficiently localize odor sources.
摘要:The characteristics of Duffing oscillator's phase trajectory in great periodic motion are analyzed,base on which,a new frequency measuring method is proposed.The simulation results indicate,the detecting signal-to-noise ratio lower limit of the proposed method is higher than that of the existed strong referential detecting method.However,the proposed method has no blind angle,and the frequency detection range is greatly enlarged.As a result,the complexity of the measuring system can be reduced on a large scale.In this method,one Duffing oscillator is capable of measuring chirp signal,which is an exploration for the application of the detection technologies based on Duffing oscillators.
摘要:With the development of satellite remote sensing technology,types and amount of remote sensing datum are increasing rapidly,which put forward higher requirements to data transmission capability.Two schemes are proposed to reduce cache capacity requirements in the buffer capacity restricted condition while designing Advanced Orbit System(AOS) encoding for data transmission.In the first one,the data transmission rate is increased.And in the second,the data interface between compression unit and the AOS encoding unit is designed synthetically and the timing of compressed data output is decentralized so as to reduce the instantaneous data rate of the AOS encoding cache entry.The buffer capacity demands in both schemes are analyzed dynamically using simulation techniques.The results show that each scheme can reduce cache capacity requirements,while the second scheme requires less buffer capacity significantly.The schemes have been applied to the data transmission system products of certain satellite engineering project and have passed flying verifications.The approaches proposed in this paper have a reference value for the follow-up low cost,high rate transmission system applications.
关键词:remote sensing satellite;high rate data transmission;buffer depth;dynamic analysis
摘要:The semantic caching can efficiently support the applications in the context of mobile computing.Most existing algorithms for query trimming suffer high time and space complexity and they can't be used in small mobile devices.To this end,we propose 20 rules and a dynamic semantic merging strategy to simplify the complexity of query trimming,which is measured by the numbers of predicates that are depicted in the description for the semantic cache or queries.Some experiments on Android system show that disjunction simplification algorithm along with Completive Merging strategy can get the best performance for simple queries among the alternatives.For complex queries,the dynamic merging strategy based on predicate simplification can balance the complexity between caches and query processing and gain a good performance in a wide range.
摘要:The mainstream lexicalized syntactic parsing models currently only utilized semantic dependency between words,and did not incorporate other semantic information such as semantic collocation and semantic category."valency"is a essential semantic feature of words,once the valency of word is determined,the collocation of the word is clear,and the sentence structure can be directly derived.Thus a syntactic parsing model combining valence structure with semantic dependency is purposed on the base of head-driven statistical syntactic parsing models.The parsing model incorporates rich semantic information including semantic dependency and semantic collocation in the decomposition and probability computation of the rules.Experiments are conducted for the refined statistical parser,it achieves 88.76% precision and 87.43% recall,F measure is improved 6.65% comparing with the head-driven parsing model introduced by Collins.
关键词:valence structure;semantic dependency;head-driven;statistical syntactic parsing model
摘要:Based on the notion of MV-algebra semantics,probability measure is set up in MV-algebra evaluation lattice and set of all propositions,and a probability truth degree of propositions in lattice-valued logic is proposed with integral.Thus pseudo-metric in set of all propositions is induced,probability logic metric space is established in lattice-valued logic,and graded reasoning is developed.In summary,approximate reasoning method in quantitative logic is expanded to lattice-valued logic,and it is feasible in graded in lattice-valued logic.
关键词:MV-algebra;lattice-valued logic;probability truth degree;probability logic metric space;approximate reasoning
摘要:Measurement matrix,whose performance can affect the compression and reconstruction of original signal,plays a key role in compressive sensing.Most of the existing measurement matrices are random ones,which have shortcomings in practical application,such as large storage capacity,low efficiency and difficulty when implemented in the hardware.Therefore,it is of important practical significance to construct deterministic measurement matrix for the promotion and application of the compressive sensing theory.In this paper,the existing construction algorithms for deterministic measurement matrix are reviewed,introduced and classified in detail.Finally the performances of all algorithms are summarized in terms of common indicators.
摘要:The multi-field coupling models of microwave reflector antennas,plate slot-array antennas and active phased array antennas have been developed in this communication,based on which the solution strategies and methods are presented to determine efficiently the effects of systematic errors,random errors and temperature etc.on the electrical performance of antennas.Some experiments are performed to validate these models and methodologies.Finally,a multidisciplinary optimization design model based on the muli-field coupling theory is derived with employing a common design vector,which can greatly reduce the repetitive works and ease the debugging of products.
摘要:In OFDM based cognitive radio(CR) systems,the large sidelobe of frequency spectrum of secondary users (SU) may cause severe interference to the primary user and the sidelobe spectrum must be suppressed.Active interference cancellation (AIC) can't null the spectrum deep enough and it is sensitive to the cyclic prefix (CP).In this paper,a non-orthogonal AIC (NAIC) technique is proposed for the sidelobe suppression of OFDM CRs.In NAIC,cancellation carriers (CCs) are inserted at the locations which are non-orthogonal to the subcarriers of the SU such that the sidelobe can be more significantly suppressed.Moreover,the windowing technique is used to reduce interference effects from non-orthogonal CCs (NAIC-win) to SU and more deeply suppress the sidelobe spectrum of SU.NAIC-win can obtain sidelobe suppression depth 50dB more than AIC with very little inter-carrier interference (ICI).Compared to EAIC-CP,NAIC-win can suppress the spectrum sidelobe deeper,increase the spectrum efficiency,and lower the computation burden with shorter CP,less CC and ICI.Theoretical analysis and numerical simulations validate that NAIC-win method significantly outperforms the conventional AIC and EAIC-CP methods in sidelobe suppression for OFDM CR systems.
摘要:In Underwater Sensor Networks (UWSNs) based on clustered architecture,cluster heads forward large number of data packets when they send data to the Sink using a single-hop or multi-hop mechanism.This mechanism can overload cluster heads,which soon become depleted of energy,causing coverage holes quickly.The LEACH-Coverage-U algorithm attempts to solve this problem,but because the cluster heads are randomly elected,this algorithm cannot solve this problem very well.In this paper,we propose a single-hop coverage-preserving routing algorithm (SCPR) for UWSNs.This algorithm first defines a metric of coverage redundancy (CR),and elects the cluster heads according to this metric.A cluster head sends the aggregated data directly to the Sink.To reduce energy consumption and achieve higher network coverage ratio,we also propose a multi-hop coverage-preserving routing algorithm (MCPR),where each cluster head prefers to select the node that has higher CR in the parent nodes as its next hop node.Finally,the aggregated data are sent to the Sink using multi-hop routing.Simulation results show that,compared to the LEACH-Coverage-U algorithm,both the SCPR and MCPR algorithms solve the problem that the cluster heads are randomly elected,improve the network coverage ratio,reduce network energy consumption and extend network lifetime.
摘要:Bugs in software are inevitable.The study on the networked software abnormal behavior propagation mechanism trigged by bugs provided the way for people to grasp the execution rule and to adopt corresponding pinning measurements.Based on the situation of abnormal behavior propagation at different granularity software entities,three factors——propagation probability,interactive frequency,and connection rate——affecting the abnormal behavior propagation were proposed,and the corresponding definition and calculation method were also investigated.The software abnormal behavior propagation process model was constructed in reference to the compartment model and individual models and three factors mentioned above,which improved model expression ability and enhanced the model competence and accuracy.Then,the abnormal behavior propagation analytical method was applied to the typical networked software system,and the results verified the correctness and feasibility of the propagation mechanism.
摘要:Based on some enterprise-level workflow authorization mechanisms,a multi-dimensional and generalizable framework for Separation of Duty is specified,and multiple labor dividing forms related to team-collaborated tasks can be restricted deeply and all-sidedly.Coverage rules among these constraints are analyzed systematically.The correctness and completeness of these rules are proved such that a basis for the automation of constraint administration is provided.As application of the rules,a detecting algorithm for redundant dynamic constraints is given.Finally,the features of this model are verified via a case study.
关键词:access control;task;role;fine-grained separation of duties;redundant constraint
摘要:In high-speed mobile environment,wireless channel shows serious time-varying characteristic.The Doppler spread caused by different Doppler frequency offset of each arrived multi-path signal worsens the performance of time-frequency synchronization algorithm.In this paper,a novel orthogonal angle domain subspace projection based time-frequency synchronization algorithm is proposed.In this algorithm,the angle resolution from multiple antennas system is utilized to simplify the Doppler spread as Doppler frequency offset on each orthogonal angle domain subspace,which can improve the performance of time-frequency synchronization algorithm.Simulation results show that the proposed algorithm suppress the effect caused by time-varying wireless channel and achieve better system performance.
摘要:Based on the similarity of chaos in human brain and Chua's circuit,we propose the methods to model human brain with Chua's circuit.With the simulation of Multisim 10®,we set voltages across resistor R1 as original signals.Also,Analogous definite integral,Lyapunov,Micro-segmented windows and P values are applied to gain data window.Then,minimum mean power of R1 (PR1min:1.56×10-4W) is discovered after tuning inductance L and nonlinear-resistor parameter R5.Furthermore,we suppose PR1minas log-base,and calculate the new indexes for each power of R1 as reference values.After compared with previous indexes of brain power consumption(BPC)in frontal lobe at cognitive states of calmness,honesty,or lying(BPC at smile as base),two-set indexes equal to each other.According to actual BPC,and with Chua's circuit corresponding to minimum mean power PR1min,we accomplish human brain model in frontal lobe at cognitive states of calmness,honesty,lying and smile.This work will provide novel thinking for nonlinear modeling of human brain.
摘要:Optimal sensor placement is important content of testability design for complicated systems,which belongs to the typical combinatorial optimization problem.Based on the improvement of the fault-sensor correlation matrix,the constraint optimization model is set up which considers the fault detection abilities of the sensors.The ergodic of chaos has been used to initialize the parameters of the particles,and the inertia weight is adjusted adaptively according to the swarm's premature convergence degree.Besides,the update of the particle's position has been redefined,then the improved discrete PSO algorithm is used to solve the optimization model.The simulation examples demonstrate that the proposed method is effective,and the optimization results can satisfy all the requests of the system,and it is a feasible approach for optimal sensor placement for complicated systems.
摘要:Evidence theory is widely used in decision-level information fusion.However,there are also some problems to be solved.For instance,how to measure the conflict between two bodies of evidence is still an open issue.Thus,a new method with good behaviors to measure conflict is proposed.Different bodies of evidence are described as a BPA matrix whose properties are studied afterwards.Considering the properties of singular values of such a BPA matrix,the smaller one is chosen to measure the conflict.The effectiveness of the proposed method is illustrated by some examples.
关键词:evidence theory;evidence conflict;singular value;basic probability assignment