The mirror method, the electric axis method and the complex conformal mapping method are the typical methods to solve electromagnetic-field problems. The image method is used only to solve the field distribution of charges near the conductor. The electrical axis method is used merely to solve the potential function between two infinite long parallel cylindrical conductors.Therefore, the conventional electric axis method is relaced by the uniform conformal mapping method in this paper, and the issues of eccentric coaxial lines and parallel pairs of cylinders are unified. This paper also presents the complex examples of the three-wire transmission line capacitor. Therefore, the unified conformal mapping method can replace the image method and electrical axis method and simplify the solution of the electromagnetic-field problem.
The Intersecting Cortical Model (ICM) is adopted for image segmentation. Because the nonlinear exponential attenuation threshold mechanism is suitable for brightness respondence but not suitable for image segmentation, we change it to the linear threshold mechanism which leads to the LT-ICM. Meanwhile, an improved 2-D Tsallis cross-entropy segmentation rule is proposed for determining the segmentation threshold and iteration time automatically. Experiments show that the proposed algorithm is better than OSTU and K-means algorithms for the texture image segmentation. Besides, our rule is more effective than the 2-D max-Shannon-entropy rule, traditional 2-D Tsallis cross-entropy rule, and min-cross-entropy rule.
In order to implement the multi-parameter estimation for the bistatic radar without any synchronization,a multi-targets signal model of the MIMO bistatic radar is set up and a joint angles-Doppler frequencies estimation algorithm for multi-targets is proposed. In the algorithm, the virtual array data of multitarget echoes in MIMO receivers is first uncorrelated, then a special matrix based on the propagation operator constructed, and finally the joint estimation of the DODs, DOAs and Doppler frequencies can be obtained by the corresponding relationships between the eigenvalue or eigenvector and the estimated parameters. The proposed algorithm does not need the multi-dimensional nonlinear peak search, and only needs an eigenvalue decomposition, and the estimated parameters of the targets can be paired automatically. The correctness and effectiveness of the proposed method are verified with computer simulation.
Considering the fact that the SDH network and technology are widely used in practice, we study the static virtual topology design problem in IP over SDH over WDM(IP/SDH/WDM) optical networks, and propose a hybrid genetic and tabu search algorithm for virtual topology designs (HGTSA-VTD). The key elements within HGTSA-VTD are designed in detail, making this hybrid algorithm have the integrated advantages of the global search for genetic algorithms and local search for tabu search algorithms. Its effectiveness in resource utility has been verified by comparing the simulation results with the results by existing approaches used formerly on the NSFnet.
To improve the robustness of the watermarking method remarkably, this paper proposes a video semantics watermarking method, which combines video semantics with video watermarking by using the high-level semantic features of stability and unvulnerability. This method looks upon mining association rules as a constrained optimization problem, uses an improved clonal selection algorithm for the fuzzy association rule mining to extract the semantics of motion and texture, and generates dynamic semantics watermark online. This method determines the shots of interest according to motion semantics adaptively, determines the I frames of interest according to texture semantics adaptively, and selects the intense movement and slow movement of the region as a region of interest according to human visual masking properties, and then embeds watermark in the IF DCT coefficients of the luminance sub-block prediction residual of the I frames of interest; it adaptively controls watermark embedding strength by use of video texture features. Experiments and analysis show that this method is robust not only to various conventional attacks, but also to re-frame, frame cropping, frame deletion and other video-specific attacks.
The multi-granularity OXC can reduce the number of switching ports of the wavelength switching plane of the ordinary OXC in the DWDM network. But there is a tight coupling between the waveband switching plane and the wavelength switching plane in ordinary multi-granularity OXC. So the switching architecture of the ordinary multi-granularity OXC is complex. A new multi-granularity OXC is proposed. In this new architecture, the wavelength switching plane and waveband switching plane are independent. Therefore the switching architecture can be simplified and the number of switching ports of the wavelength switching plane can be reduced effectively. In order to increase the utilization of the waveband, a traffic grooming algorithm for this new architecture is proposed. Simulation results show that the traffic grooming algorithm can effectively decease the blocking possibility.
An inversion channel electron mobility model of the 4H-SiC n-MOSFET is presented according to the physical mechanism. The quasi-2D Coulomb scattering mobility model based on the first principles takes account of the screening effect by carriers and temperature on Coulomb scattering, in which no empirical parameters are included. This can be directly embedded into the 2-D device simulator. The numerical relationship between the proportionality constant of the surface roughness scattering mobility model and the surface roughness degree is deduced. Correctness of the model is confirmed by comparison between simulated and experimental results. The simulated results show that Coulomb scattering plays a dominant role near the interface. Surface roughness scattering becomes more obvious at a higher gate voltage and it is the dominant scattering mechanism to limit the mobility in the inversion layer for the higher doped SiC MOSFET.
Most existing differential evolution algorithms for the Constrained Optimization Problem(COP) use the penalty function method to handle constrains, which depends strongly on the penalty parameter. So, this paper transforms the COP into two-objective multi-objective optimization by taking constraints as an objective function. Based on the concept of Pareto, the grades of individuals in population are prescribed so as to determine their selection probability in the process of “survival of the fittest”. In addition, when the algorithm gets into a local optimum, an infeasible solution replacing mechanism is also given to improve the search capability. The results of the 13 Standard tests show that compared to the Evolutionary Algorithm based on Homomorphous Maps (EAHM), Constraint Handling Differential Evolution (CHDE), Evolutionary Strategies based on Stochastic Ranking (ESSR) and Artificial Immune Response Constrained Evolutionary Strategy (AIRCES), the proposed algorithm has certain advantages in convergence speed and solution accuracy.
AA drive circuit for driving an external power MOSFET is designed for active power factor correction (APFC), which consists of a level-shift circuit and a single totem-pole output stage. The level-shift circuit adopts the current mirror structure so that it can reduce the power dissipation by adjusting the bias current. The dead time in the single totem-pole output stage can also decrease the power dissipation. Clamping the control signal VGS for the high voltage PMOS between 6V and 11V not only makes low consumption but also saves the layout area. Based on the 0.4μm BCD process, simulation results show that the better circuit power is 7.34mW and that the layout area can be saved by 15% when VDD is 14V and switch frequency is 75kHz.
For the error diagnosis problem in object-oriented software maintenance, this paper proposes a novel alias analysis technique. By extending the control flow graph of a program, exception handling structures are analyzed. Based on the extended control flow graph, the process of alias analysis is divided into two phases, flow sensitive intra-class analysis and requirements based inter-class analysis, to achieve a complete alias analysis of the whole system, and effectively improves the efficiency of error diagnosis for programs which contain exception handling structures.
A robust successive MMSE (SMMSE) precoding scheme is proposed in the context of the multiple user multiple input multiple output (MU-MIMO) downlink (DL) transmission system under the condition of imperfect channel state information (CSI) at the transmitter side. The proposed algorithm enables us to use multiple antennas at the receiving terminal so that it is possible to jointly process the received signal at the receiving terminial, with the eigen-beamforming technique belonging to the joint processing method. Simulation results show that the robust SMMSE algorithm aided eigen-beamforing technique can effectively improve the system performance compared to the one without considering the impact of the channel estimation error, and even completely remove the error floor.
Considering the long sweep period of the Frequency Modulated Continuous Wave (FMCW) SAR, it becomes difficult to maintain the linearity of frequency. For lack of the satisfactory correction methods, a modified frequency non-linearity correction algorithm is presented. Furthermore, the Doppler frequency shift introduced by the platform's continuous motion is analyzed in detail. For the signals after frequency non-linearity correction, the concept of instantaneous frequency is used by introducing the parameter, i.e., bistatic angle, to deduce the signal analytical expression in the two dimensional wave-number domain. Based on this, a modified Range Migration Algorithm(RMA) with frequency non-linearity correction is proposed which is aimed at the Tandem Bistatic FMCW SAR. Simulation data processing verifies the correctness of the analysis and the effectiveness of the algorithms, including frequency non linearity correction and imaging processing.
The performance of amplify-forward (AF) and decode-forward (DF) is analyzed in the two basic single relaying modes for a single-relay system in the Rayleigh fading environment. This paper investigates the impact of the relay's position on the system capacity and probability of error for both AF and DF schemes. The position of the relay is crucial to the performance. The result shows that depending on transmission power, channel noise and path loss the best performance is achieved when the relay is at an equal distance from the sender and the destination in AF or the relay is closer to the source around in DF.
The finite-difference time-domain(FDTD) method is extended to two dimensional anisotropic dielectric dispersive media—magnetized plasma by using Shift Operator(SO). It is proved that when the incident plane wave vector is perpendicular to the axis of the 2-dimensional axial magnetized plasma cylinder, the electromagnetic wave can be decomposed into Transverse Electric(TE) and Transverse Magnetic(TM) waves, with the electromagnetic scattering characters of the TM wave having no relation with the external DC magnetic field but the electromagnetic scattering characters of the TE wave having close relations with the electromagnetic parameters in magnetized plasma. The physical reasons have been analyzed. The FDTD with shift operator (SO-FDTD) formulations in this case is derived based on the constitutive relation for the magnetized plasma and shift operator. For electromagnetic scattering of the Von Karman shaped conducting cylinder coated with anisotropic magnetized plasma, the effect of coating thickness and plasma parameters on the bistatic Radar Cross Section(RCS) is investigated, with some numerical results obtained. The results indicate that appropriate plasma coating may efficiently reduce the RCS of a metallic target.
Most of the speckle filtering methods can not preserve the scattering characteristics of the PolSAR image because of their being based on the statistic model. To solve this problem, here we propose a speckle filtering method to preserve polarimetric scattering characteristics. Firstly, by extracting scattering characteristics from the PolSAR image, we mark different sorts with different markers; then, referring to the different markers, an adaptive slipping window is used to compute the filtering parameter; finally, measured data are applied in the filtering experiment using the filtering parameter. From the analysis, we have proved the effectiveness of our method.
In order to extend the swath of the SAR (Synthetic Aperture Radar) image to match the scene, the missile-borne ScanSAR wide-swath mode is studied in this paper. Firstly, the geometric model and the range equation are built, secondly the spatial variation of the linear range cell migration and the Doppler centroid are analyzed, thirdly the modified Range-Doppler algorithm and its computation load are analyzed, and lastly the geometric correction expressions are deduced. Simulation shows the effectiveness of the imaging mode and algorithms proposed.
In the process of filtering, the filtering accuracy can be improved if the state constraints are used in an effective manner. The nonlinear constrained function can be linearized by the Taylor series expansion. However, if the Jacobian matrix of the nonlinear constrained function is nonexistent, this method will not work anymore. Moreover, the moving horizon estimation (MHE) algorithm needs a heavy computational burden in this condition. So a perfect measurement method is proposed based on the unscented transform to solve this problem. Furthermore, in order to reduce the negative effect from the base point error, the nonlinear constraints can be treated as measurements with different noise covariance. The noise covariance shrinks in the measurement update stage, and the constrained conditions are enhanced step by step. The state estimation error is improved after some iterations. Simulation results show that the proposed algorithm can obtain a higher filtering accuracy, and that its computational time is 1/27 of that of the moving horizon estimation algorithm even if the window size is 2.
It is a practical and feasible way to acquire multibaseline interferometric data by utilizing repeat passes of a single-baseline InSAR system. Due to the high coherence, the single-pass pair of SAR images generated by a single-baseline InSAR can be coregistered using conventional methods. However, there exist extreme difficulties in the coregistration operation for multiple-pass SAR images due to the serious temporal and spatial baseline decorrelation, especially in the case of large effective cross-track baseline lengths. Fortunately, the repeat-pass interferograms obtained by a single-baseline InSAR correspond to the identical terrain, i.e., Digital Elevation Model (DEM). Hence the interferograms contain coherent information, even if the SAR images are decorrelated completely. Based on this point, we propose an innovative strategy to coregister the repeat-pass interferograms instead of the SAR images. Computer simulation results show that the proposed method has the ability to align multiple-pass interferograms of a single-baseline InSAR system accurately.
In multi-target tracking, aiming at the data association problem and the curse of dimensionality that arises due to the increased size of the state-space associated with multiple targets, a novel algorithm based on the parallel Gaussian particle filter (parallel-GPF) is proposed. By assuming that the states of the targets conditioned on the past measurements are mutually independent, the algorithm is amenable for multi-channel parallel implementation. With the Markov chain Monte Carlo (MCMC) approach the marginal association probabilities can be evaluated. Then, the filtering distributions of each target are computed independently with GPF. And, Quasi Monte Carlo (QMC) integration is utilized for approximating the prediction and update distributions. Finally, the proposed method is applied to passive multi-sensor multi-target tracking. Simulation results show that the proposed algorithm can obtain a better tracking performance than JPDAF, MCMCDAF and MC-JPDAF.
For the problem of the forward-looking blind zone on the missile-borne SAR, this paper presents a squint RD imaging algorithm in a missile trajectory. This algorithm requires a curve locus to receive target imaging, which causes the radar antenna to be directed to the target at a squint degree that changes the forward-looking into a squint mode. Considering the handicap by the complex trajectory equation under the curve locus, we first obtain the trajectory equation in high order, and takes RCMC(Range Cell Migration Correction) in the time-doppler domain, then derive the two-dimensional frequency domain expression by series reversion and finally receive the imaging results by a two-dimensional matched filter. Simulation results verifies its validity.
For the low system degrees of freedom of the Space-Time Multiple-Beam (STMB) approach and the general sidelobe canceller structure of the multistage Wiener filter (MWF), a dimension-reducing space-time adaptive processing (STAP) based on the STMB approach for the airborne radar is proposed. Firstly, to reduce the degrees of freedom of the clutter, the space-time received data are pre-filtered by using the STMB approach; secondly, the output data after preprocessing are processed by the MWF to optimize the weight vector, which can further reduce both computational complexity and sample support requirement. Theoretical analysis and computer simulation results illustrate that the method can achieve a robust performance without a large loss in the presence of large random amplitude and phase errors existing in the array elements. Experimental results by using the measured data demonstrate the effectiveness of the proposed method.
Aiming at the motion compensation for the modulated stepped frequency radar signal echoed by a moving target, we develop a new method combining envelope correlation and the least burst error rule for estimating the velocity. First, the coarse velocity is measured by using envelope correlation, then the refined velocity is estimated by utilizing the least burst error rule based on the coarse velocity, which avoids the problem of the local minimum existing in the least burst error method. Simulation results indicate that the method can accomplish accurate estimation of target motion parameters with good anti-noise performance.
The decentralized state feedback control problem is studied for a nonlinear interconnected systems with a time-varying delay in both states and inputs which are composed of a number of Takagi-Sugeno (T-S) fuzzy bilinear subsystems with interconnections. Based on the Lyapunov criterion and the parallel distribution compensation scheme, the delay-dependent stabilization sufficient conditions are derived for the whole close-loop fuzzy interconnected systems, which have less conservativeness than that by the existing methods. The corresponding decentralized fuzzy controller design is converted into a convex optimization problem with linear matrix inequality (LMI) constraints. Finally, a simulation example shows the effectiveness of the proposed approach.
Two design methods for nonbinary rate-compatible LDPC(RC-LDPC)codes are proposed. And a novel irregular bit-wise shortening algorithm based on the algebraic properties of their binary images is proposed, with which we can obtain improved performance and extra flexibility; a symbol-wise puncturing algorithm for binary LDPC codes is extended to nonbinary LDPC codes. With these two algorithms, we obtain a family of nonbinary RC-LDPC codes with the code-rate range from 1/10 to 5/6. Simulation results demonstrate that gains of 0.1~1.1dB are obtained by proposed methods.
Based on inequality constraint, numerical optimization is used in the traditional quadratic programming method to compensate the passivity violations, with the required simulation time being quite large. A novel quadratic programming method to enforce the passivity of macromodels is presented, which can greatly reduce the simulation time. This method is based on the quadratic programming method with an equality constraint. The associated optimization problem can be solved by the method of Lagrange multipliers. The Krylov subspace method is adopted to solve the resulting linear system, which can fully utilize the sparsity of the problem. The computational time required for this method is significantly reduced. A frequency selection strategy for passivity violations is also introduced. Several examples show that the solving time required for the presented method is less than 1/10 that of the traditional quadratic programming method.
Cognitive radio (CR) is an intelligent wireless communication system, and the design of a cognitive engine is one of the key issues. The cognitive engine can utilize the reasoning and learning methods to achieve the perception, adaptation and learning ability. This paper presents a design scheme for the cognitive engine based on the RBF neural network (RBF-NN) to adapt varied wireless environment according to experiential knowledge. A CR simulation cognitive engine is realized using the 802.11a simulation platform by collecting data and training the NN to learn channel characteristics. Under the premise of studying the sensation information, CR reconfiguration is realized according to the user's needs. Simulation results show that the proposed cognitive engine can effectively realize the study and reconfiguration function, and that the prediction accuracy is higher than that of the BP NN model by 10%~30%.
In order to improve the accuracy of vertical handover and to reduce the “Ping-Pong effect”, a difference threshold based fuzzy handover (DTFH) algorithm is proposed for heterogeneous networks. The handover trigger and handover decision are considered for DTFH. The signal to noise ratio (SNR) gap between the target network and the current network is obtained by the mobile station (MS), and the SNR gap between the two networks is compared with a predefined trigger threshold for handover trigger. Then based on the method of fuzzy logic, the MS gets the fuzzy decision values of the target network and the current network, and the gap between the fuzzy decision values of these two networks is compared with a predefined decision threshold for handover decision. In simulation, a heterogeneous network comprising LTE and WiMAX is used. Compared with traditional handover algorithms, the simulation results show that the DTFH algorithm can reduce the “Ping-Pong effect”, increase the throughput of MS, and improve the performance of vertical handover.
Traditional software reliability growth models based on the black-box approach are inappropriate to model the reliability of component-based software. How to analyze and predict software reliability from components' reliabilities is concerned. Considering that the component developer and user may be different and taking into account the assembling methods of component-based software, a reliability prediction model of component-based embedded software is proposed. Using this model, the reliability of the system is a function of the stochastic properties of the component failure behavior, the architecture structure of the program and the utilization of these components by the user. The main ideas of this model are: firstly, the software architecture is modeled as a discrete time Markov chain, and complex structures such as parallel, redundant and subroutine calls are transformed from the architecture view to the state view, secondly, the developer of the component gives each interface's reliability of the component and the user provides the usage profile from the component level and interface level, on the basis of which we can calculate the component's reliability in different use scenarios.
In order to enable the mobile network to provide anonymity services and ensure the confidentiality of the user's identity, whereabouts and other information, we put up an anonymous authentication scheme, which solves the problem of user's anonymity not only in roaming network, but also in adscription network. The scheme has not only a small amount of communication, but also does not need to change the key every time. The scheme is based on the knowledge proof, and uses the direct anonymous attestation protocol theory together with encrypting transfer and signature validation for its implementation. Theoretical analysis and experimental results demonstrate that users do not need to go to the home agent to authenticate the user's identity in the roaming network, and that there is no need to map out its true identity in the adscription network. It implements user's anonymity in the whole network effectively.
The number of antennas at the base station side in a multi-user distributed MIMO system is much larger than that of a centralized one. Accordingly, antenna selection at the base station side becomes one of the key technologies in such a system. However, traditional algorithms are too complex to be realized. A low complexity receive antenna selection algorithm based on parameter optimization is proposed to deal with the complexity problem in the uplink of multi-user distributed MIMO systems. In order to reduce the complexity, an elimination method with capacity computation replaced by iterative parameter optimization computation is introduced in the algorithm. Both complexity analysis and simulation results show that the computation complexity is largely reduced while the capacity performance is still close to that of the optimal algorithm.
Since the equal power allocation (EPA) between the source node and the relay node in the decode-forward (DF) cooperative communication system results in resource inefficiency, a power allocation algorithm based on outage probability analysis is proposed. This algorithm aims to reduce the average outage probability of the system, in which the upper and lower bound expressions for the system average outage probability of the model are derived according to the average signal to noise ratios of the upper link channels of and the channels between users. The power allocation factor that leads to a lower system average outage probability is obtained by the golden section method. Simulation results show that, all conditions being the same, this power allocation scheme can achieve a better performance than the equal power allocation (EPA) scheme.
A simple, yet effective decoding algorithm for LDPC (low-density parity-check) codes over GF(q) is presented. By taking advantages of the first-term Taylor's series expansion to approximate the correction term of the Jacobian logarithm used in LLR-SPA(log-likelihood ratio sum-product algorithm), we propose an algorithm which significantly simplifies the check node update computation of the optimal LLR-SPA. Compared to the offset min-sum algorithm, the proposed algorithm achieves a gain of about 0.2dB at the BER of 10-4. Moreover, unlike the offset min-sum algorithm, parameters of this algorithm are independent of the order of the Galois Filed.
A new hierarchical modulation based sphere detection algorithm (HBSD) is presented for the MIMO (multi-input multi-output) spatial multiplex system with high order modulation. In the algorithm, the constellations of MQAM (M-Quadrature Amplitude Modulation) are divided into many regions by the hierarchical data in MQAM modulation. The candidate regions of the constellations are determined by the hard decision data of the layers with great weight factors in hierarchical modulation. The constellations are selected as the reserved constellations in the candidate regions using the minmum distance rule. SD (Sphere detection) is used to find the optimum transmitted vectors by searching for the reserved constellations in the candidate regions of all detection layers. Simulation results show that the performance of the proposed HBSD algorithm is better than that of the MSD (Multi-SD) algorithm, with its complexity being about one fifth that of the MSD algorithm.
Since the major application scenarios of the DTN (delay/disruption tolerant network) are extremely particular, routing protocols designed for the traditional networks are not suitable for it. In order to raise the transfer efficiency while reducing the network delay and network cost in the DTN, a routing algorithm based on the Markov Decision, named MRDDTN (Markov Route Decision in the DTN), is proposed. The MRDDTN is self-learning and can solve problems in the complex network environment characterized by very long delay paths and frequent network partitions. Through simulative analysis, the MRDDTN is found to be able to achieve a better performance than Epidemic and PROPHET routing algorithms under the same network conditions.
Packet loss is a primary element to impair the quality of network speech, and the impairment extent is related not only to the number of lost packets, but also to their content feature. This paper proposes a no-reference quality assessment model for networked speech based on analysis of the content feature of packets. This model first estimates the content feature of lost packets by resorting to the voice active detection technology and the level of non-lost frame. Based on the content feature analysis, the proposed model calculates the speech packet loss rate and then predicts the quality of networked speech. Experimental results show that high prediction accuracy can be obtained by the proposed model. Compared with the speech quality evaluation model in G.1070, the average performance gain of the proposed model is raised 8.4% in terms of relativity to subjective results.
A new relay scheme based on analog network coding and joint iterative detection & decoding method is proposed for the multiple-access full-duplex relay system to further improve the bit error rate and frame error rate. The information packets from different users are superimposed after interleaving and modulation at the relay node, and then transmitted to the destination node. At the destination, the joint iterative detection and decoding approach is utilized to obtain source information. A higher coding gain can be obtained by fully exploiting information redundancy from analog network coding. Simulation results show that the proposed scheme can achieve about 5dB improvement on the SNR compared with iterative decoding scheme based on code superposition in the case of the ideal channel condition between source and relay node.
In order to select the relay node and improve the system throughput, a new multi-user cooperative communication protocol is proposed based on the opportunistic communication and network coding. On the basis of outage probability, each user decides whether to help forward its partner's information as a relay or not. The relay node processes its own information and the partner's information by the use of network coding, and transmits it to the receiver. The receiver can obtain several transmitters' information using coherent detection. Compared with traditional cooperative protocols, the proposed protocol can improve the average per user throughput and the network throughput in the quasi-static Rayleigh fading channel. The probability of system outage is reduced with fewer time resources.
The complexity of the iterative Demapping algorithm in Bit-Interleaved Coded Modulation systems is related to the product of the order of modulation and the constellation points number, so that it is not suitable for real-time applications. To reduce its complexity, we give a shortcut calculation based on the Hard Decision Feedback for the BICM system using LDPC codes. A soft information calculation algorithm based on the decision boundary is used in the first iteration, where no feedback information is available to be used. Simulation result shows that the BWE performance of the algorithm proposed in the AWGN channel has only a distance of 0.04~0.06dB from the soft information iteration algorithm. On the other hand, the calculation complexity of this algorithm is only related to the order of the modulation, which is more suitable for application.
A novel method for image retrieval based on interest points is presented. The interest points are detected in the scale space. Then the image is divided into fan-shaped sub-regions of equal area according to the distribution of the interest points. Local features representing the spatial distribution information on the interest points are extracted to describe the image, and they are also robust to the image's rotation, scale and translation. In the relevant feedback, images are regarded as multiple-instance bags consisting of the local domain of the interest points in every fan-shaped sub-region. Labeled images chosen by the user are generated corresponding positive and negative bags, and the multiple-instance learning algorithm is employed to obtain the target concept reflecting the query image semantics. The method can reduce the ambiguity of the user query. Experimental results based on the Core image database show that our method improves the average retrieval precision by 7 percent or more, compared with other interest points based retrieval methods.
To improve the Bluetooth-based WPAN (Wireless Personal Area Network) throughput, the system models were built, and the chaotic algorithm was designed to improve the WPAN throughput. The system models includes the WPAN network model and the data collision model. The data collision crushes the data packet and makes the data resending, and then the function of WPAN data throughput is deduced with the aforementioned conclusion. It can be proved that the number of piconets and the hamming correlation ability of the FH (Frequency Hopping) algorithm decide the WPAN throughput. The new chaotic FH algorithm based on the Chebyshev map was built. It has a better hamming correlation ability than the Bluetooth FH algorithm and can improve the WPAN throughput. The computational simulation and actual test were done. The test used the ARM1510 SoC (System on Chip)+nRF2401 radio chip to build the WPAN. Simulation and test results prove that the new chaotic FH algorithm can increase the WPAN maximum throughput about twice, in other words it can improve the number of piconets in the WPAN about twice with the same throughput.