摘要:An integrated RF PLL frequency synthesizer is presented.It integrates VCO,dual-modulus prescaler,PFD,Charge-pump,various digital counters,control logic and the series interface with the base-band processor into a single chip.Also the selection of internal VCO or external VCO and power control are implemented to adapt to various applications.The frequency synthesizer has been implemented in 0.25μm CMOS process.The measured results show that the locked range is 1.82GHz-1.96GHz when the internal VCO is selected,the phase noise could reach -119.25dBc/Hz at 25MHz offset from the carrier 1.924GHz.The analog part uses a 2.7V power supply and the consumed current is about 48mA.
摘要:For a class of strict feedback nonlinear system,a nonlinear adaptive controller was first investigated.The design procedure used backstepping method based on DBFNN (Direction Basis Neural Network).In each step of backstepping a virtual controller was designed by introducing a suitable Lyapunov function.In the last step the singular-free real controller was synthesized.The tuning law of NN weights was derived from the selected integral Lyapunov function.So the stability of the closed loop can be guaranteed. Then the proposed scheme was applied to design an excitation controller for one power system with unknown disturbance and unknown parameters.The simulation shows the validity of the proposed method.
摘要:We discuss the problems of solving linear systems with multiple right-hand sides while using PEEC model to extract 3-D parasitic interconnect inductance and resistance in VLSI circuits.Based on the seed projection method EGCR (Extended Generalized Conjugate Residual)[5],we propose a modified method MEGCR to simplify the computation of orthogonal procedure through limitedly enlarging the search space of non-seed right-hand sides.It accelerates the speed of solving systems with right-hand sides.Summarizing the technique of solving linear systems with multiple right-hand sides and multipole acceleration,our program runs faster than FastHenry[1] tens' times with comparable accuracy when the number of right-hand sides is about 30.
关键词:3-D parasitic inductance;PEEC;multipole acceleration;linear systems with multiple right-hand sides
摘要:Routing based on a minimum connected dominating set (MCDS) in ad hoc wireless networks is a promising approach,where the search space for a route is reduced to nodes in the set (also called gateway nodes).This paper introduces MWMCDS,a simple and efficient heuristic algorithm for calculating minimum connected dominating set with maximal weight in the topology graph G of an Ad hoc wireless network.The maximality of the weight-based choice of gateway nodes guarantees that the most suitable nodes have been chosen for the role of gateway nodes so that they can properly coordinate all the other nodes in the network.As a result,it can keep stability of the MCDS and provide a highly effective communication base for broadcast and routing operation in the whole network.Simulation resluts show that the proposed algorithm can ensure the maximality of sum of CDS' weight and the minimality of CDS' size.So the scheme can be potentially used in designing efficient routing algorithms based on a MCDS.
摘要:The canonical re-coding and sliding window techniques are often used in computation of scalar multiplication k·P on elliptic curves for reducing the average number of required operation.In this paper,scalar multiplication with canonical re-coding and sliding window techniques is analyzed by modeling the window partition process of canonical re-coding expression of k as Markov-chain,the average performance of scalar multiplication under different parameters are given and the optimal window sizes are computed.Finally,the comparison shows that scalar multiplication with canonical re-coding and sliding window techniques requires 10.32~17.32% fewer operations than m-ary method,and 4.53~8.40% fewer operations than simple sliding window method.
YUAN Zhi-peng, LIU Hong-gang, LIU Xun-chun, WU De-xin
Vol. 32, Issue 11, Pages: 1782-1784(2004)
摘要:A 10Gb/s driver IC for optic modulator has been implemented using our 4-inch InGaP/GaAs HBT process.The driver IC operates up to 10Gb/s with an output voltage swing of 3Vpp and 34.2ps(20 to 80%) of tr and 37.8ps(20 to 80%) of tf.and input return losses of -12.3dB at 10GHz,which make the driver suitable for 10Gb/s optic telecommunication systems(SONET OC-192,SDH STM-64).A single power supply VEE is -5.2V,and the power consumption is 1.3W.The die size is 2.01×1.38mm2.
YIN Qing-bo, ZHANG RU-bo, LI Xue-yao, WANG Hui-qiang
Vol. 32, Issue 11, Pages: 1785-1788(2004)
摘要:A new method for anomaly intrusion detection is proposed based on dynamic Markov model.At first,behavioral features are extracted from the privileged processes,and then the Markov model is founded dynamically based on the features.The state sequences of dynamic Markov model are analyzed to infer the state probability,which is used to classify the normal or abnormal behavior.Because Markov model is constructed dynamically,it can extract the relationships of local behavioral features of the privileged processes adequately.When the training sets are limited,the method predicts exactly.The experiments show this method is simple,effective and efficient,and can be used in practice to monitor the computer system in real time.
PAN Wei, ZHANG Xiao-xia, LUO Bin, DENG Guo, LI Xiao-feng, CHEN Jian-guo
Vol. 32, Issue 11, Pages: 1789-1792(2004)
摘要:A new method of VCSELs for controlling chaos is studied,which is based on the parameter switching modulation.The chaos the system is controlled to its fixed point and nP periodic orbits by using the spontaneous emission factor.proves the validity and robustness of this proposed control method.It is shown that high-order harmonic distortion is minimized with large spontaneous emission factor.The quantitative relation among the stability,multi-period or the position of the bifurcation and the parameter switching are given using the bifurcation diagram,the temporal waveforms and the phase space portrait.
LV Jin, SHI Yi, PU Lin, YANG Hong-guan, YANG Zheng, ZHENG You-dou
Vol. 32, Issue 11, Pages: 1793-1795(2004)
摘要:The time characteristics of the composite quantum dots based MOSFET memory is simulated with the Monte Carlo method in quasiclassical approximation.It indicates that the retention time could be improved evidently owing to the stepwise compound potential barrier.As an example,the time characteristics of N channel Ge/Si hetero-nanocrystal based MOSFET memory is investigated and the retention time could be as long as several years,at the same time,the writing and erasing time can be in the order of μs and ns,respectively.Hence the conflict between high-speed programming and long retention could be resolved satisfying.
REN Feng-yuan, LIN Chuang, HUANG Xiao-meng, LIU Wei-dong
Vol. 32, Issue 11, Pages: 1796-1800(2004)
摘要:Active Queue Management (AQM) is an enhancement mechanism to end-to-end congestion control,which can maintain smaller queuing delay and higher throughput by purposefully dropping the packets at intermediate nodes.Almost all the existing AQM schemes follow the probability dropping mechanism originated from Random Early Detection (RED).It is essentially a decision process with aid of information about network congestion that AQM decides whether or not to drop an incoming packet,thus the probability discrimination should not be only one way.We firstly summarize the properties of an ideal AQM scheme,and then develop a novel framework for AQM scheme based on the two-category classifier,which is considerably simple and extensible.A two-dimensional two-category classifier (TCC) for AQM is designed using the Fisher Linear Discriminate approach.The simulation results show that TCC outperforms other well-known AQM schemes in the integrated performance,namely TCC is an effective,prompt and robust algorithm.Additionally,TCC scheme requires few CPU cycles,which makes it deploy in high-speed routers simply.
LIN Ya-ping, WANG Lei, CHEN Yu, ZHONG Jin, CHEN Zhi-ping, TONG Tiao-sheng
Vol. 32, Issue 11, Pages: 1801-1805(2004)
摘要:Sensor networks have severe energy constraints,redundant low-rate data,and many-to-one flows.The traditional end-to-end routing algorithms are usually central and are not adapted for sensor networks.A distributed data-centric clustering hierarchical algorithm is proposed,which uses the energy-core concept to perform aggregation of data in the network and to reduce the amount of information that must be transmitted to the sink.Simulations show that,compared with the traditional end-to-end algorithm,this algorithm can significantly reduce energy dissipation.Compared with the data-centric algorithm without clustering,the data-centric clustering hierarchical algorithm offers better performance gains in complexity and scalability,while still being energy efficient.
LAN Xu-guang, ZHENG Nan-ning, MEI Kui-zhi, LIU Yue-hu
Vol. 32, Issue 11, Pages: 1806-1809(2004)
摘要:A real-time parallel array architecture is proposed to perform the forward and inverse discrete wavelet transform (DWT) by using lifting scheme for the filters recommended in JPEG2000.It consists of two row processors and one column processor.The signals are processed in parallel way.Multiplications are substituted for shift-add operations.The whole architecture is optimized in the pipeline design way to speed up and achieve higher hardware utilization.It performs a decomposition in approximately N2/2 clock cycles for an N×N image.A compact and independent IP core based on the architecture for JPEG2000 VLSI implementation has been demonstrated in FPGA.
摘要:Location management,which consists of location updating and location query,is a challenging topic in mobile communication.The location query of the basic location management scheme (Basic Scheme for short) used in personal communications services (PCS) networks in many countries, requires to page a mobile in all the cells within a location area (LA) simultaneously.This scheme is likely to waste the system's channel resource and processing resource enormously,since the LA consists of many cells whereas the mobile only resides in one.This paper proposed a new location management scheme,Cell Layered Scheme,which grouped all the cells of each LA into layers and made terminal paging one layer after another in probability order.Both the probability of the mobile residing in each layer and the average number of cells searched by the proposed scheme are derived on the assumption that the residence time of a mobile residing in each cell is a generally distributed random variable.It is proved that the cost of the Cell Layered Scheme is no more than that of the Basic Scheme.
关键词:location management;mobility management;personal communication networks (PCN);vector markov process
XIONG Zhi-hui, LI Si-kun, CHEN Ji-hua, WANG Hai-li, BIAN Ji-nian
Vol. 32, Issue 11, Pages: 1815-1819(2004)
摘要:In order to attain separation of function from structure and separation of computation from communication,we present hierarchical Platform-Based Design (Hi-PBD) method for SoC system design.This method divides SoC system design flow into 3 levels (i.e.system model level,virtual components level and real components level).Hi-PBD performs 2 mappings among the 3 design levels via design planning and virtual-real synthesis.Our method not only reuses the design templates of the 3 levels,but also reuses the mapping results,which increases reusing efficiency.Besides,Hi-PBD boosts up design flexibility through supporting revision at all the 3 levels,and ensures final design target to satisfy performance requirements through performance constraints transmission.Experiments indicate that Hi-PBD method improves SoC high level design efficiency by 30%-40%,and this method achieves platform template reuse ratio by 75%-90%.
ZHANG Zhi-wei, GUO Chang-guo, CAI Jun-ya, WU Quan-yuan
Vol. 32, Issue 11, Pages: 1820-1823(2004)
摘要:Asynchronous mechanism is absolutely necessary in the construction of large scale distributed systems.Although being a popular support platform for distributed applications,CORBA doesn't solve the asynchronous communication mechanism problem very well and this limits its application to some application domains.How to provide asynchronous mechanism in CORBA and provide asynchronous communication support for applications at upper layer have become the research hotspot.This paper presents a CORBA asynchronous messaging model Star-Async.The model supports reply processing,exception handling and asynchronous code generation by means of ReplyHandler object,exception encapsulation object and modification of abstract syntax tree accordingly.We have implemented Star-Async model on StarBus,which is our own distributed object-oriented middleware platform implementation.The implementation results of Star-Async and its initial applicationsshow that Star-Async model provides a reference model for the implementation of asynchronous mechanism in CORBA.
关键词:CORBA asynchronous messaging;asynchronous mechanism;callback model;polling model
摘要:Premature convergence is the fatal shortcoming of traditional evolutionary programming.Based on the analysis of premature convergence of traditional evolutionary programming,a novel Multi-subgroup Competition Evolutionary Programming (MCEP) algorithm is proposed.In the algorithm,evolutions of subgroups are parallely performed with the different mutation strategies.The mutation capability of an individual is determined by the competition among subgroups.The individual,together with information,is exchanged while subgroup is reorganized.The simulations based on benchmarks confirm that this algorithm is better than classical evolutionary programming algorithm in the aspects of global optimization,convergence speed and the robustness.
摘要:Aiming at normal PWM low efficiency at light loads,the author proposed a novel high efficiency modulation mode named pulse skip modulation mode PSM.This paper establishes the large signal model of the novel PSM switching converter mode,which includes large analytical model,discrete model and phase plane model.Basing on its analytical model,discrete model and phase portrait,the transient characteristics of the open and closed PSM converter are given in this paper.Compared with PWM converter,PSM converter has quicker response and better interference rejection besides high efficiency.
关键词:PSM converter;large signal model;analytical model;discrete model;phase portrait technique;transient response
摘要:The three basic problems of two-dimensional (2-D) hidden Markov models (HMMs) are studied,including probability evaluation,optimal states and parameter estimation.By using the idea that the sequences of states on columns or rows of a 2-D HMM can be seen as states of a 1-D HMM,several new analytic formulae for solving these three problems are theoretically derived and further demonstrated by computer simulation.
摘要:Traditional Stochastic High-Level Petri Nets (SHLPNs) can efficiently simplify system models by folding more than one homogeneous subsystem into one subsystem, and obviously reduce state space size by grouping more than one marking with the same token distribution into one marking.The grouped marking is called as Compound Markings (CMs).However,this method can only be applied to the systems consisting of homogeneous subsystems.This paper extends the traditional method to solve this problem that traditional SHLPNs are only limited to homogeneous systems, by proposing asymmetrical transition-firing predication and extended compound marking concept.The new method can model and analyze inhomogeneous systems exactly.Furthermore,it retains the SHLPNs advantages of simplifying models and reducing state space size.
LONG Bai-tao, SUN Li-feng, CHEN Wen-ping, ZHONG Yu-zhuo
Vol. 32, Issue 11, Pages: 1844-1848(2004)
摘要:This paper investigates the impact of congestion on the performance of spanning trees in Application-layer Multicast (ALM) environments and proposes using redundant links between adjacent levels to enhance the stability of spanning trees. We analyze this approach by developing a simple statistical link model which describes congestion occurring in spanning trees and by introducing a dependency-degree factor which models the dependence between adjacent links.The numerical simulation results indicate that the statistical link model well describes the congestion occurring in spanning trees for ALM,and that even in case of high congestion probability,the redundant links scheme can significantly enhance spanning trees' resilience to congestion at a low cost of additional network delay.The proposed statistical link model is generic and can be utilized not only to further understand the dynamics in ALM,but also to design algorithms of congestion control for ALM.
关键词:application-layer multicast;statistical link model;inter-level redundant link
LI Yun, LIU Zong-tian, CHEN Ling, XU Xiao-hua, CHENG Wei
Vol. 32, Issue 11, Pages: 1849-1854(2004)
摘要:Since the completeness of concept lattice,the time complexity of building concept lattice is a factor restricting the application of formal concept analysis.Based on the horizontal and vertical combination in formal contexts,this paper defines the independent or consistent contexts and lattices in attribute field;and also defines the horizontal addition operation between contexts or concepts and the horizontal union operation between concept lattices.In addition,we prove that the concept lattice of subcontexts horizontally combined is isomorphic to the horizontal union of sublattices of these subcontexts.Using the inherent general-special relation between concepts in sublattice,the horizontal union algorithm of multiple concept lattices to construct the concept lattice is also presented.Experimental results show that the time complexity of this algorithm is much better than that of other construction algorithm of concept lattice from whole formal context.Evidently,our algorithm is very suitable for constructing concept lattice in parallel and distributed system.
关键词:concept lattice;formal context;sublattice;subcontext;horizontal union
XIONG Yan, CHEN Huan-huan, MIAO Fu-you, WANG Xing-fu
Vol. 32, Issue 11, Pages: 1855-1858(2004)
摘要:Based on quantum mutation, a quantum genetic algorithm (QGA) to solve combinatorial optimization problem is proposed.It has good features of genetic quantum algorithm (GQA) and traditional genetic algorithm,which can obtain the best solution with one chromosome in a short time.Comparing experiments have been conducted on a typical combinatorial optimization problem——0/1 knapsack problem.Experimental results have shown that the proposed quantum genetic algorithm is superior to genetic quantum algorithm and traditional genetic algorithm on performance.
摘要:Based on a novel all-phase FIR adaptive filter,combination of overlapping idea with LMS algorithm in transform domain is proposed.Overlapping windowed adaptive filtering algorithms(WO-TLMS) in DFT、DCT and DST transform domain are presented.Compared with the conventional LMS algorithm in transform domain,WO-TLMS algorithm owns lower mean square steady error and more rapid convergence speed.Convergence property of WO-TLMS is theoretically analyzed and the superiority to TLMS is verified experimentally.
摘要:Feature vector method for fast fractal image encoding is considered as one of the most innovative and promising approaches,but it suffers from several drawbacks,especially high dimensionality of feature vectors.Thus,an alternative feature method to reduce fractal encoding time is proposed.As one of its applications,cross trace-based fast fractal algorithm is presented,where the cross trace is a newly-defined feature of an image block.The proposed algorithm converts the range-domain block matching problem to the neighborhood search problem in the sense of cross trace.A simulation on popular 256×256 Lena image shows that,depending on the search window size,the proposed algorithm not only can achieve the speed-up of over 3 times with the same PSNR (peak signal-to-noise ratio) as the baseline fractal algorithm with the full search,but also can obtain the speed-up of 100 times or more at the cost of tolerable degradation of the decoded image quality.
摘要:Computing the coefficients is the key to obtain the all-terminal reliability of communication by using reliability polynomial,and computing some coefficients exactly is NP-hard.After studying the relation between connected subnet numbers and the cutest of the network,this paper presents a method for computing the cutest numbers and a formula of the coefficient bounds of reliability polynomial.Finally,when the edge life is exponentially distributed,the paper computes the bounds of all-terminal reliability of a SDH network to demonstrate the method.
GAO Feng, LIU Qi-zhong, SHAN Run-hong, XIAO Liang-yong
Vol. 32, Issue 11, Pages: 1871-1874(2004)
摘要:This paper investigates the benefit of mutual coupling compensation in a novel smart antenna circular array at 3G-communication system frequency.Mutual coupling of antenna arrays and the compensation of steering vector and array manifold are analysed via method of moment (MoM).Based on it,the paper implements the compensation of downlink beamforming and uplink DOA algorithm.Finally,experiments in antenna array prove the efficiency of mutual coupling compensation.
关键词:smart antenna;circular array;mutual coupling compensation;coaxial collinear;method of moment
摘要:The experimental results of wideband antenna radiating and receiving ultra fast electrical pulse that was generated by GaAs photoconductive switch triggered by femo-second laser pulse is reported.From receiving antenna,ultra fast electrical pulse of 200 ps rise time and 500 ps pulse width was obtained,the repetition rate of the pulse was about 82 MHz and frequency bandwidth was more than 6.0 GHz.
摘要:A fast method for wide-band flat-gain fiber Raman amplifier pump configuration is provided in this paper.It can give the pump wavelength and power level in Least Mean Square (LMS) sense,given input parameters like,fiber parameters,signal power level,target gain profile and number of allowed pumps.Based on LMS approach,this method combines the Genetic Algorithm (GA) to determine pump wavelength and convex quadratic programming with linear constrained conditions to determine pump power levels,so that it speeds up the searching speed.
摘要:Digital signature is an important security mechanism which is applied on the network security environment.The Smart card or Token is an optimal hardware to realize digital signature.How to implement digital signature on the Smart card or Token is a key issue.Based on contact smart card standards and series standards of PKCS(Public Key Cryptographic Standard),this paper implemented the RSA(a kind of public key asymmetric algorithm) digital signature、identification、message encryption、keys distribution,especially emphasized on analyzing some key questions about how to realize RSA on the Smart card practically,and present several method scheme.
摘要:ECC uses much shorter key than RSA does in the case of the same security.Although the mathematical theory of ECC has become perfect, understanding and implementing its algorithms is much more difficult than those of RSA.It often needs special ECC chip to speed the algorithm implementing.This paper introduces a kind of entire ECC algorithms we have designed.To raise the efficiency of the ECC implementing,the algorithms for point multiplying and square overplus judging have been optimized,so that the whole ECC algorithms can be implemented through software.
摘要:This paper presents a multi-model fuzzy-control algorithm for non-linear Group Decision Support System-GDSS.The algorithm can control the change of model parameters according to model -matching degree,and calls different models flexibly to solve specific problems so that the solution speed and precision can be improved.Simulation shows that this algorithm is simple,efficient and feasible.
摘要:Lifting factorization (LF) was one of the latest advancements of study on fast DWT implementations.Because of exponential size of the factorization set,known algorithms could not find in reasonable time all stable (to say nothing of optimal) LFs of a given DWT with long filters.In this paper,a so-called constant GCD (Greatest Common Divisor) factorization approach to FIR polyphase matrix was given,which contracted considerably the factorization space.Furthermore,the problem of how to evaluate a certain LF was investigated in terms of both numerical stability and computation complexity.Consequently,an algorithm called OLFA (Optimal LF Algorithm) was readily available.All of the theoretical results were constructively proven.Experimental data show that OLFA obtains considerable improvements in solution quality,computation time and application range over the existing algorithms,thus makes LF a tool of great generality and practicability for fast DWT implementations.
关键词:DWT;optimal lifting factorization;Euclidean algorithm;stability of factorization
摘要:A competing classification method is presented to extract Web information.The method uses similarity between information fragments and samples as competing ability.It classifies fragments and filters out noise information through competition of fragments for template slots.It needs far less tagged samples than those using rules to extract information.Experiments show that the method keeps high precision of information extraction without any feature clues provided by users.Therefore it is adaptive.The competing classification method is also robust in dealing with data sources having missing items and items of various orders.
摘要:A time-delay recurrent neural network (TDRNN) model is proposed.TDRNN has a simple structure but far more "depth" and "resolution ratio" in memory by introducing the time-delay and recurrent mechanism.A TDRNN controller is utilized for controlling ultrasonic motors.A dynamic recurrent back propagation algorithm is developed.To guarantee the fast convergence of the proposed neural network model,the optimal adaptive learning rates are derived in the sense of discrete-type Lyapunov stability.Numeral experiments for controlling speeds of ultrasonic motors show that the TDRNN has good effectiveness in the identification and control for dynamic systems.
CHEN Min-si, YAO Su-ying, ZHAO Yi-qiang, ZHANG Sheng-cai, LI Shu-rong, XU Jiang-tao, WANG Tian-sheng
Vol. 32, Issue 11, Pages: 1922-1925(2004)
摘要:The system design and implementation methods of timing control circuit for a new CMOS image sensor are proposed.Using dual-sampling pixel architecture,the CMOS image sensor can achieve both high frame rate and wide dynamic range.Two efficient timing control methods(one named "roll exposure mode",and the other "parallel exposure mode") are both proved right and feasible,by analyzing their theoretical operation principle and the results from function and timing simulation and FPGA verification.
摘要:We propose an approach for the formal modeling and analysis of security properties of cryptographic protocols.Our methodology is based on the idea of backward state analysis and the reachability analysis for Petri nets.By performing the state analysis,we can find out potential insecure states during the run of protocols and determine if these insecure states are reachable by using the state reachability analysis.An example shows that this approach is efficacious.
摘要:This paper deals with the method to raise the front to back(F/B) ratio on the pattern for the circularly polariged microsyrip antenna patch in small size.Here the EBG construction on the patch opposite plane is adopted to be the ground and the continual tooth around the circumambient edges of the EBG ground are made,so that the F/B is increased for the patch ground in small size.The electric parameters of the designed patch are measured.The experimental results shows that this method is effective.
关键词:microstrip patch;EBG construction;pattern;front to back ratio
摘要:The noise correlation matrixes of cascade representation,impedance representation and conductance representation of a two-port network were derived from basic circuit and noise definitions.The noise correlation matrixes of two linked two-port networks were also gotten from basic definitions.Then the noise parameters and S parameters of a parallel feedback low noise amplifier were derived using these correlation matrixes.A 1~7 GHz two-stage MMIC LNA was designed on the basis of above formulas and fabricated using 0.5μm GaAs PHEMT processes.After measurement,it is found that from frequency 1.0 GHz to 7.0 GHz,the LNA achieves a power gain of 20 dB as well as a good flatness of only ±0.75 dB,a maximum noise figure of 2.5 dB,input and output VSWRs of better than 2.0 and the output P1dB power of 15 dBm.The design is verified by the measurement results of the MMIC LNA.