ISSN 
2002 Vol. 31, No. 2
2002, 31(2): 111-114.
Abstract:
The influence of tunable transceiver and wavelength converter to the performance of survivable WDM network under the condition of dynamic traffic is studied in this paper. By simulation in NSFNET(the U.S. National Science Foundation backbone network), CERNET(China Education and Research network) and MESH-TORUS network, we analyze the blocking performance of wavelength converter and tunable transceiver in survivable WDM network, the result shows that although the wavelength converter and tunable transceiver can improve the blocking performance, the performance gain of wavelength converter is much better than that of tunable transceiver under dynamic traffic in survivable WDM network.
The influence of tunable transceiver and wavelength converter to the performance of survivable WDM network under the condition of dynamic traffic is studied in this paper. By simulation in NSFNET(the U.S. National Science Foundation backbone network), CERNET(China Education and Research network) and MESH-TORUS network, we analyze the blocking performance of wavelength converter and tunable transceiver in survivable WDM network, the result shows that although the wavelength converter and tunable transceiver can improve the blocking performance, the performance gain of wavelength converter is much better than that of tunable transceiver under dynamic traffic in survivable WDM network.
2002, 31(2): 115-119.
Abstract:
A call admission control policy in wideband multimedia CDMA system is proposed in this paper. Different interference thresholds are set up for new calls and handoff calls. When new or handoff call arrives, interference threshold for new call and handoff call are firstly determined, then intra-cell and inter-cell interference are calculated, and the total interference is got. The total interference is compared with interference threshold, if the total interference is less than the threshold, the call is accepted. The policy improves the system performance with lower handoff call dropping probability.
A call admission control policy in wideband multimedia CDMA system is proposed in this paper. Different interference thresholds are set up for new calls and handoff calls. When new or handoff call arrives, interference threshold for new call and handoff call are firstly determined, then intra-cell and inter-cell interference are calculated, and the total interference is got. The total interference is compared with interference threshold, if the total interference is less than the threshold, the call is accepted. The policy improves the system performance with lower handoff call dropping probability.
2002, 31(2): 120-124.
Abstract:
Radar signal digital pulse compression processing can be realized mostly by time-domain and frequency-domain methods. The special-purpose DSP chips have been used for the digital pulse compression processing systems widely. At the same time, the general-purpose DSP of high performance will make systems design more agile. The digital pulse compression time-domain processing system and frequency-domain processing system with the special-purpose DSP chips have been developed. On the basis of the real-time testing results obtained and through computer simulating, we made relatively overall analysis and comparison between these two methods, using the current technology of the general-purpose DSP.
Radar signal digital pulse compression processing can be realized mostly by time-domain and frequency-domain methods. The special-purpose DSP chips have been used for the digital pulse compression processing systems widely. At the same time, the general-purpose DSP of high performance will make systems design more agile. The digital pulse compression time-domain processing system and frequency-domain processing system with the special-purpose DSP chips have been developed. On the basis of the real-time testing results obtained and through computer simulating, we made relatively overall analysis and comparison between these two methods, using the current technology of the general-purpose DSP.
2002, 31(2): 125-130.
Abstract:
The robustness of a watermark is determined by the watermark strength. The adjustment of the watermark intensity is made manually now, it is suitable for a few images. In this paper, a method that can automate adjustment of watermark intensity based on the measurement of image texture is proposed, it is suitable for a larger number of images. The results of our experiments show that it is practicable to predict the watermark strength to be applied to an image based on its texture.
The robustness of a watermark is determined by the watermark strength. The adjustment of the watermark intensity is made manually now, it is suitable for a few images. In this paper, a method that can automate adjustment of watermark intensity based on the measurement of image texture is proposed, it is suitable for a larger number of images. The results of our experiments show that it is practicable to predict the watermark strength to be applied to an image based on its texture.
2002, 31(2): 131-135.
Abstract:
An interlace/progressive transforming algorithm which is to get high quality progressive frame picture from interlace field picture is proposed. In this algorithm, motion estimation technology is used to find out the motion vector between the two neighboring field pictures. Then in the process of line interpolation, the motion vector is used to adjust the first field picture's location. The motion blur, which is caused by image slip, can be eliminated. The progressive picture that is produced by the algorithm possesses high vertical resolution. In the algorithm, we design a special motion estimation approach to fit in with this application.
An interlace/progressive transforming algorithm which is to get high quality progressive frame picture from interlace field picture is proposed. In this algorithm, motion estimation technology is used to find out the motion vector between the two neighboring field pictures. Then in the process of line interpolation, the motion vector is used to adjust the first field picture's location. The motion blur, which is caused by image slip, can be eliminated. The progressive picture that is produced by the algorithm possesses high vertical resolution. In the algorithm, we design a special motion estimation approach to fit in with this application.
2002, 31(2): 136-140.
Abstract:
The circuits 4-Value TTL NOT gate and NAND gate, which designed using bipolar transistor, have better I/O character, stronger energy of load and noise immunity, their noise margin may reach ±0.6 V. Especially, the circuit of NAND gate my be totally made by NPN transistor, and the circuit structure is simple, easily is complicated by Integrate Circuit. The circuits introduced in this paper is useful, may be used in basic cell circuit of 4-Value Digital System, Artificial Neural Network and Fuzzy Logical System.
The circuits 4-Value TTL NOT gate and NAND gate, which designed using bipolar transistor, have better I/O character, stronger energy of load and noise immunity, their noise margin may reach ±0.6 V. Especially, the circuit of NAND gate my be totally made by NPN transistor, and the circuit structure is simple, easily is complicated by Integrate Circuit. The circuits introduced in this paper is useful, may be used in basic cell circuit of 4-Value Digital System, Artificial Neural Network and Fuzzy Logical System.
2002, 31(2): 141-144.
Abstract:
Nonlinear control is the purpose, direct analysis is the route, and the concept of "inversion" of dynamic system is the kernel of this theory. The author applies the inverse system method to rigid body dynamics system and designs a control of the Euler dynamic equations, so that the angular velocity of a rigid body keeps track of the dictate in the coordinate of motion. The progressive tracking of the motion of a rigid body is realized by controlling outside moment of a rotor.
Nonlinear control is the purpose, direct analysis is the route, and the concept of "inversion" of dynamic system is the kernel of this theory. The author applies the inverse system method to rigid body dynamics system and designs a control of the Euler dynamic equations, so that the angular velocity of a rigid body keeps track of the dictate in the coordinate of motion. The progressive tracking of the motion of a rigid body is realized by controlling outside moment of a rotor.
2002, 31(2): 145-150.
Abstract:
The application of the workflow management in IT fields is discussed generally. The basic concepts of the workflow are studied in detail. After put forwarding the workflow management reference model, its components and characteristics are probed into. Then, according the reference model, the three kinds of interoperability among (between) the workflow engines are referred. Through analyzing four general design modes of the workflow interoperability, their characteristics and features are discussed and compared. Then, the classification of the interoperability of the workflow is elucidated. The two ways to realize the interoperability of the different workflow system are addressed. Finally, the shortcomings and the trends in the future of the current workflow interoperability are also touched on.
The application of the workflow management in IT fields is discussed generally. The basic concepts of the workflow are studied in detail. After put forwarding the workflow management reference model, its components and characteristics are probed into. Then, according the reference model, the three kinds of interoperability among (between) the workflow engines are referred. Through analyzing four general design modes of the workflow interoperability, their characteristics and features are discussed and compared. Then, the classification of the interoperability of the workflow is elucidated. The two ways to realize the interoperability of the different workflow system are addressed. Finally, the shortcomings and the trends in the future of the current workflow interoperability are also touched on.
2002, 31(2): 151-155.
Abstract:
The probability diagnosis idea is presented in this paper; A system model suitable for fault diagnosis purpose, which takes into account the dependencies between the different objects in the Local-Area Networks (LAN) is introduced. For fault diagnosis problem is a NP-C problem, a heuristic algorithm is developed. The algorithm overcomes the disadvantages of existing fault diagnosis algorithms like high computing complex and ignoring fault dependency. It is proved to be effective by our simulation results. And it can generate acceptable results with good performances.
The probability diagnosis idea is presented in this paper; A system model suitable for fault diagnosis purpose, which takes into account the dependencies between the different objects in the Local-Area Networks (LAN) is introduced. For fault diagnosis problem is a NP-C problem, a heuristic algorithm is developed. The algorithm overcomes the disadvantages of existing fault diagnosis algorithms like high computing complex and ignoring fault dependency. It is proved to be effective by our simulation results. And it can generate acceptable results with good performances.
2002, 31(2): 156-158.
Abstract:
In this paper, the structure of LogP model is introduced, the time of network delay is more than processor, such as delay speeding, interface cost, the cost of TCP protocol, the time of message passing,the effects of is bigger than processor about parallel efficiency based on LogP in light load. As the results, it is more signification in order to reduce communication delay and improve parallel efficiency that choose highness rapidity's communications equipment than choose processor of high speed in network parallel environment.
In this paper, the structure of LogP model is introduced, the time of network delay is more than processor, such as delay speeding, interface cost, the cost of TCP protocol, the time of message passing,the effects of is bigger than processor about parallel efficiency based on LogP in light load. As the results, it is more signification in order to reduce communication delay and improve parallel efficiency that choose highness rapidity's communications equipment than choose processor of high speed in network parallel environment.
2002, 31(2): 159-162.
Abstract:
With neural networks, the main difficult in improving the model generalization capability is controlling the complexity of the model. This paper investigates a Bayesian neural network learning for nonparametric regression. Prior knowledge about the model parameters can be incorporated within Bayesian inference and combined with training data to control complexity of different parts of the model. A Markov chain Monte Carlo algorithm is used to optimize model control parameters and obtain the predictive distribution. We show that the complexity of the models adapts to the complexity of the data and produces good results on five noisy test functions in two dimension. The performance and advantage of this approach are compared with conventional neural network methods.
With neural networks, the main difficult in improving the model generalization capability is controlling the complexity of the model. This paper investigates a Bayesian neural network learning for nonparametric regression. Prior knowledge about the model parameters can be incorporated within Bayesian inference and combined with training data to control complexity of different parts of the model. A Markov chain Monte Carlo algorithm is used to optimize model control parameters and obtain the predictive distribution. We show that the complexity of the models adapts to the complexity of the data and produces good results on five noisy test functions in two dimension. The performance and advantage of this approach are compared with conventional neural network methods.
2002, 31(2): 163-167.
Abstract:
In this paper, we study the Hopf bifurcation phenomena of a neural network with a continuous time delay. Using the average time delay as bifurcation parameter, we have proved that the model undergoes a sequence of Hopf bifurcations in both the strong kernel and weak kernel cases. Stability criteria for the bifurcating periodic solutions are derived in the neural network with weak kernel. Some numerical examples and the computer simulation results are also presented to justify the theoretical results.
In this paper, we study the Hopf bifurcation phenomena of a neural network with a continuous time delay. Using the average time delay as bifurcation parameter, we have proved that the model undergoes a sequence of Hopf bifurcations in both the strong kernel and weak kernel cases. Stability criteria for the bifurcating periodic solutions are derived in the neural network with weak kernel. Some numerical examples and the computer simulation results are also presented to justify the theoretical results.
2002, 31(2): 168-172.
Abstract:
Importance of Intelligent Transportation System(ITS) and its trend are discussed. The networking architecture of ITS is put forwarded and analyzed. Then, the middleware used in ITS is touched on and its key place in ITS is also pointed out. After the design of the intelligent transportation system message queuing middleware(ITS-MQ), its constructs and components are illustrated. Then, this paper focus on the analysis of the process of ITS-MQ and then its features are also refereed. In order to verify and formalize the protocol used in ITS-MQ, we adopt the Colored Petri Net(CPN) to describe it. And the result shows that this technology is very useful to do this. The result also proves that ITS-MQ middleware is applicable for ITS.
Importance of Intelligent Transportation System(ITS) and its trend are discussed. The networking architecture of ITS is put forwarded and analyzed. Then, the middleware used in ITS is touched on and its key place in ITS is also pointed out. After the design of the intelligent transportation system message queuing middleware(ITS-MQ), its constructs and components are illustrated. Then, this paper focus on the analysis of the process of ITS-MQ and then its features are also refereed. In order to verify and formalize the protocol used in ITS-MQ, we adopt the Colored Petri Net(CPN) to describe it. And the result shows that this technology is very useful to do this. The result also proves that ITS-MQ middleware is applicable for ITS.
2002, 31(2): 173-175.
Abstract:
In this paper, the structure of several models is introduced, such as Maste/Slave, Single program Multiple Data, Data Pipelining, Divide and Conquer, Speculative Parallelism, and the working is expatiated, both characteristic and performance analysis of parallel models is also given, A quantity of application programming is analyzes and stated. The result shows parallel models influence directly correctness and efficiency of parallel programming. Both master-slave and SPMD is best programming model because of that is best of all in common use models with the all and the one performance.
In this paper, the structure of several models is introduced, such as Maste/Slave, Single program Multiple Data, Data Pipelining, Divide and Conquer, Speculative Parallelism, and the working is expatiated, both characteristic and performance analysis of parallel models is also given, A quantity of application programming is analyzes and stated. The result shows parallel models influence directly correctness and efficiency of parallel programming. Both master-slave and SPMD is best programming model because of that is best of all in common use models with the all and the one performance.
2002, 31(2): 176-179.
Abstract:
This paper discusses the method to improve data acquisition rates of logic analyzer, presents a new technique to increase the equivalent data sampling rates of logic analyzer and to position the high-speed data view window in the data stream. By making use of this technology, it can take a small cost to improve performances of logic analyzer and improve the flexibility of the design.
This paper discusses the method to improve data acquisition rates of logic analyzer, presents a new technique to increase the equivalent data sampling rates of logic analyzer and to position the high-speed data view window in the data stream. By making use of this technology, it can take a small cost to improve performances of logic analyzer and improve the flexibility of the design.
2002, 31(2): 180-184.
Abstract:
A simple survey of routing and wavelength assignment problem in WDM all-optical network under static network state is presented in this paper. We develop a novel classification for the RWA based on some of their main features, and analysis the performance of all algorithms, following with some conclusions that may be useful for a network maker to choose a reasonable RWA algorithm. According to the algorithm performance and the requirements of network optimization, some problems that need further research in RWA field are indicated.
A simple survey of routing and wavelength assignment problem in WDM all-optical network under static network state is presented in this paper. We develop a novel classification for the RWA based on some of their main features, and analysis the performance of all algorithms, following with some conclusions that may be useful for a network maker to choose a reasonable RWA algorithm. According to the algorithm performance and the requirements of network optimization, some problems that need further research in RWA field are indicated.
2002, 31(2): 185-187.
Abstract:
The technique of wave recording by software triggering is introduced, which is used in the monitoring system of equipment state. Through alternately storing successive triggering datas in two data blocks, it can achieve about-trigger wave recording for successive triggering. By means of labelling the priorities of the triggering sources and stopping or shielding the triggering of low priority by the triggering of high priority, it solves the problem of triggering overlap of multi software triggering sources.
The technique of wave recording by software triggering is introduced, which is used in the monitoring system of equipment state. Through alternately storing successive triggering datas in two data blocks, it can achieve about-trigger wave recording for successive triggering. By means of labelling the priorities of the triggering sources and stopping or shielding the triggering of low priority by the triggering of high priority, it solves the problem of triggering overlap of multi software triggering sources.
2002, 31(2): 188-191.
Abstract:
Intrusion Detection is a newly developed area of network security. The main issue in this area is how to pick-up and analyze the information which contains abnormal network behavior characteristic. In this paper, basing on the research of CIDF and the implement strategy of intrusion Detection System, we design a component-based Intrusion Detection System, which has good distribute and scalable ability. It combine the network-based IDS and host-based IDS into a system, and provide detection, report and respond together.
Intrusion Detection is a newly developed area of network security. The main issue in this area is how to pick-up and analyze the information which contains abnormal network behavior characteristic. In this paper, basing on the research of CIDF and the implement strategy of intrusion Detection System, we design a component-based Intrusion Detection System, which has good distribute and scalable ability. It combine the network-based IDS and host-based IDS into a system, and provide detection, report and respond together.
2002, 31(2): 192-195.
Abstract:
This paper put up detailed analysis and summarize about common three bilinear transformations on stability, frequently characteristic and dynamic characteristic in the control system. Three transformations have the same feature on stability analysis, and that have the distinct difference on frequently characteristic analysis and dynamic characteristic analysis. The paper also discusses the right applied conditions about the existent problems of three transformations. The material example is illuminated. It is beneficial to apply rightly bilinear transformation.
This paper put up detailed analysis and summarize about common three bilinear transformations on stability, frequently characteristic and dynamic characteristic in the control system. Three transformations have the same feature on stability analysis, and that have the distinct difference on frequently characteristic analysis and dynamic characteristic analysis. The paper also discusses the right applied conditions about the existent problems of three transformations. The material example is illuminated. It is beneficial to apply rightly bilinear transformation.
2002, 31(2): 196-199.
Abstract:
This paper mainly discusses miniature isolated solid-state encapsulation technology of high-temperature polysilicon pressure sensor, including static electricity bonding,stainless steel diaphragm selection and rippled design,laser welding,silicon oil infilling, isolation and other technologies used in sensor packaging. By adopting stainless steel diaphragm and high-temperature silicon oil as isolation materials, not only the encapsulation diameter of sensor is as small as 15 mm and under 1 mA drive,its full range output is 72 mV and zero stability is 0.48% F.S/mon,but also the reliability of the sensor is improved and its application is widely broadened.
This paper mainly discusses miniature isolated solid-state encapsulation technology of high-temperature polysilicon pressure sensor, including static electricity bonding,stainless steel diaphragm selection and rippled design,laser welding,silicon oil infilling, isolation and other technologies used in sensor packaging. By adopting stainless steel diaphragm and high-temperature silicon oil as isolation materials, not only the encapsulation diameter of sensor is as small as 15 mm and under 1 mA drive,its full range output is 72 mV and zero stability is 0.48% F.S/mon,but also the reliability of the sensor is improved and its application is widely broadened.
2002, 31(2): 200-203.
Abstract:
This paper analyzes the influence of artificial factors on the product reliability, proposes an optimization model of combined system reliability, and quantitatively analyzes the increasing degree of artificial reliability factors in the case of parallel connection. Furthermore, this paper confirms the effectiveness of the optimization model of combined system reliability by the calculation of simulated value.
This paper analyzes the influence of artificial factors on the product reliability, proposes an optimization model of combined system reliability, and quantitatively analyzes the increasing degree of artificial reliability factors in the case of parallel connection. Furthermore, this paper confirms the effectiveness of the optimization model of combined system reliability by the calculation of simulated value.
2002, 31(2): 204-206.
Abstract:
Using growth elements as cluster variables, analysis of economic growth in Hierachical Cluster Analysis is given in this paper. The different effect on economic growth is probed by the cluster. And the economic growth of Building Material Industry from 1991 to 1998 was analyzed by SPSS, it showed there was a more great effect.
Using growth elements as cluster variables, analysis of economic growth in Hierachical Cluster Analysis is given in this paper. The different effect on economic growth is probed by the cluster. And the economic growth of Building Material Industry from 1991 to 1998 was analyzed by SPSS, it showed there was a more great effect.
