ISSN 
2003 Vol. 32, No. 4
2003, 32(4): 349-353.
Abstract:
In this paper, the coverage and capacity in CDMA Systems with multi-user detection technique are analyzed. A genetic algorithm based multi-user detection technique is presented. With the computer simulation, it is shown that the new multi-user detection technique has better performance than the conventional multi-user detection technique in terms of both the resistance on the multiple access interference and the near-far effects.
In this paper, the coverage and capacity in CDMA Systems with multi-user detection technique are analyzed. A genetic algorithm based multi-user detection technique is presented. With the computer simulation, it is shown that the new multi-user detection technique has better performance than the conventional multi-user detection technique in terms of both the resistance on the multiple access interference and the near-far effects.
2003, 32(4): 354-357,379.
Abstract:
In WCDMA, how to use the power control to decrease the transmittion power and lessen the system interference to upper the system capacity on the cond ition of assure the Qos users required is one of the keypoints. The main work of the paper is to analyze and summarize the power control in WCDMA, including the inner loop power control, outer loop power control, uplink power control, downl ink power control, coming up with one outer-loop power power control algorithm
In WCDMA, how to use the power control to decrease the transmittion power and lessen the system interference to upper the system capacity on the cond ition of assure the Qos users required is one of the keypoints. The main work of the paper is to analyze and summarize the power control in WCDMA, including the inner loop power control, outer loop power control, uplink power control, downl ink power control, coming up with one outer-loop power power control algorithm
2003, 32(4): 358-361.
Abstract:
Location management scheme is very important in mobile communication system. During developing the CDMA2000 system, it is found that location management scheme of CDMA2000 core network works inefficiently. In this paper, an optimized scheme is proposed based on the service model of the current CDMA2000 system. Not changing the system structure of CDMA2000, this optimized scheme can work efficiently. At the same time, this scheme can be easily realized only via software alteration in the current CDMA2000 system.
Location management scheme is very important in mobile communication system. During developing the CDMA2000 system, it is found that location management scheme of CDMA2000 core network works inefficiently. In this paper, an optimized scheme is proposed based on the service model of the current CDMA2000 system. Not changing the system structure of CDMA2000, this optimized scheme can work efficiently. At the same time, this scheme can be easily realized only via software alteration in the current CDMA2000 system.
2003, 32(4): 362-366.
Abstract:
G.729 speech codec (encoder and decoder) algorithm and the TMS320VC549 fixed-point DSP (Digital Signal Processing) chip are briefly described in this paper. We put our emphasis on the hardware and software design of the low bit rate speech codec based on TMS320VC549 chip. Also, some key techniques for the real-time implementation of the complex algorithm on fixed-point DSP (Digital Signal Processor) are discussed. The results show that the real-time implementation of G.729 speech coder on fixed-point DSP can produce 8kb/s low bit rate, lower algorithm delay and toll-quality speech.
G.729 speech codec (encoder and decoder) algorithm and the TMS320VC549 fixed-point DSP (Digital Signal Processing) chip are briefly described in this paper. We put our emphasis on the hardware and software design of the low bit rate speech codec based on TMS320VC549 chip. Also, some key techniques for the real-time implementation of the complex algorithm on fixed-point DSP (Digital Signal Processor) are discussed. The results show that the real-time implementation of G.729 speech coder on fixed-point DSP can produce 8kb/s low bit rate, lower algorithm delay and toll-quality speech.
2003, 32(4): 367-370.
Abstract:
This paper introduces the architecture of GSM location system and its implementation scheme. It describes the infrared communication interface protocol of this demo system, and brings forward the status transform diagram used by the protocol. It mainly analyses the theories and methods of the location algorithm used by the demo system, and builds its channel model. Meanwhile, it analyses the testing results in the micro-cellular condition. With this it shows that this scheme accomplishes the mobile telephone relative position calculating using wireless cellular system, and will more perfectly acquire the absolute position of the mobile phone, so eventually complete the location tracing.
This paper introduces the architecture of GSM location system and its implementation scheme. It describes the infrared communication interface protocol of this demo system, and brings forward the status transform diagram used by the protocol. It mainly analyses the theories and methods of the location algorithm used by the demo system, and builds its channel model. Meanwhile, it analyses the testing results in the micro-cellular condition. With this it shows that this scheme accomplishes the mobile telephone relative position calculating using wireless cellular system, and will more perfectly acquire the absolute position of the mobile phone, so eventually complete the location tracing.
2003, 32(4): 371-374.
Abstract:
Based on the GPS original technology and the electronic MAP, a new practical system solution to the vehicle information service system in GSM networks by using SMS call center is presented. The design method for the GPS/GSM system includes the structure of the regional E-MAP, the organization of the car's terminal, the protocol of the isolated parts and a low-cost protocol which can improve the work efficiency and accelerate corporation information course.
Based on the GPS original technology and the electronic MAP, a new practical system solution to the vehicle information service system in GSM networks by using SMS call center is presented. The design method for the GPS/GSM system includes the structure of the regional E-MAP, the organization of the car's terminal, the protocol of the isolated parts and a low-cost protocol which can improve the work efficiency and accelerate corporation information course.
2003, 32(4): 375-379.
Abstract:
The satellite's attitude, the earth rotation and the cluster-satellite rotation will affect the baseline between two micro-satellite in the cluster satellite system. In this paper, the height-measure error influenced by satellite's attitude will be analyzed especially. Theoretical analysis and simulation results demonstrate that 0.1° attitude error will bring on 0.1m baseline error and 20.42 m height-measure error.
The satellite's attitude, the earth rotation and the cluster-satellite rotation will affect the baseline between two micro-satellite in the cluster satellite system. In this paper, the height-measure error influenced by satellite's attitude will be analyzed especially. Theoretical analysis and simulation results demonstrate that 0.1° attitude error will bring on 0.1m baseline error and 20.42 m height-measure error.
2003, 32(4): 380-384,402.
Abstract:
Reconstruction of formation conductivity distribution for 6FF40 measurement information using an inverse method was described in this paper. Based on the internal and external electrical field integral equation, the inversion equation is derived. Then, the Numerical ModeMatching is used to solve the logging response of the 6FF40. Numerical simulations show that satisfactory inversion result of the complicated formation can be obtain from the incomplete measurement of the 6ff40. In addition, prior information, the radial and vertical resolution of the 6FF40 is also investigated.
Reconstruction of formation conductivity distribution for 6FF40 measurement information using an inverse method was described in this paper. Based on the internal and external electrical field integral equation, the inversion equation is derived. Then, the Numerical ModeMatching is used to solve the logging response of the 6FF40. Numerical simulations show that satisfactory inversion result of the complicated formation can be obtain from the incomplete measurement of the 6ff40. In addition, prior information, the radial and vertical resolution of the 6FF40 is also investigated.
2003, 32(4): 385-388.
Abstract:
The traditional FDTD discretization of Maxwell equations in which all cells are equal is proved not to be efficient in many cases. In order to conform to fine scale geometric details of interconnects without increasing the overall mesh size and without losing computational accuracy, graded mesh technique combined with perfectly electric conducting boundary in 2D FDTD algorithm is proposed in this paper. Numerical experiments verify its accuracy and efficiency.
The traditional FDTD discretization of Maxwell equations in which all cells are equal is proved not to be efficient in many cases. In order to conform to fine scale geometric details of interconnects without increasing the overall mesh size and without losing computational accuracy, graded mesh technique combined with perfectly electric conducting boundary in 2D FDTD algorithm is proposed in this paper. Numerical experiments verify its accuracy and efficiency.
2003, 32(4): 389-394.
Abstract:
A numerical method for calculating the reflection coefficient of the impregnated absorbent honeycomb with plane wave normal incident to its broad face is presented. By introducing a concept of impedance for the honeycomb wall and by the assistance of the periodic Green's function, the Electric Field Integral Equation for the infinite honeycomb surface is established and the unknown surface current is solved by the method of moment. Numerical results are obtained for various heights, different absorbent parameters and various impregnated thickness within the frequency band 8~12 GHz. It shows the proposed method is of high efficiency and can be used in the honeycomb fabrication and application of structural absorbers.
A numerical method for calculating the reflection coefficient of the impregnated absorbent honeycomb with plane wave normal incident to its broad face is presented. By introducing a concept of impedance for the honeycomb wall and by the assistance of the periodic Green's function, the Electric Field Integral Equation for the infinite honeycomb surface is established and the unknown surface current is solved by the method of moment. Numerical results are obtained for various heights, different absorbent parameters and various impregnated thickness within the frequency band 8~12 GHz. It shows the proposed method is of high efficiency and can be used in the honeycomb fabrication and application of structural absorbers.
2003, 32(4): 395-398.
Abstract:
In this paper, Based on the weighted GS algorithm, as an example, a 16-phase level binary optical element designed for transforming the input Gaussian beam into nearly flat-top, The size of weight's influence to the binary optical element is numerically simulated, at the same time, we designed a 16-phase level binary optical element with the best weight, and compared with the results abstained by other weights, the simulation results show that with the increase of the weight, the averages square difference reduced, but the diffraction efficiency also reduced.
In this paper, Based on the weighted GS algorithm, as an example, a 16-phase level binary optical element designed for transforming the input Gaussian beam into nearly flat-top, The size of weight's influence to the binary optical element is numerically simulated, at the same time, we designed a 16-phase level binary optical element with the best weight, and compared with the results abstained by other weights, the simulation results show that with the increase of the weight, the averages square difference reduced, but the diffraction efficiency also reduced.
2003, 32(4): 399-402.
Abstract:
The design and calculation elements of the single anode magnetron injection gun for gyrotron is researched. with 8 mm and the third-harmoric gyrotron of the casing by the permanent magnet for examples. The Numerical calculation and optimization is carried out by use of an self-redact electron trajectory code, The space charge effects and the velocity spreads is discussed. Lastly, The optimal parameters with electrode shapes and electronic trajectory was obtained. The single anode magnetron injection gun was employed experiments and researches for 8 mm and the third-harmoric gyrotron of the casing by the permanent magnet. The results was all right.
The design and calculation elements of the single anode magnetron injection gun for gyrotron is researched. with 8 mm and the third-harmoric gyrotron of the casing by the permanent magnet for examples. The Numerical calculation and optimization is carried out by use of an self-redact electron trajectory code, The space charge effects and the velocity spreads is discussed. Lastly, The optimal parameters with electrode shapes and electronic trajectory was obtained. The single anode magnetron injection gun was employed experiments and researches for 8 mm and the third-harmoric gyrotron of the casing by the permanent magnet. The results was all right.
2003, 32(4): 403-408.
Abstract:
By expanding the surface currents on the helix tape in a series of chebyshev polynomials, simulating the helical thickness with vacuum layer, equaling the rods with multiple azimuthally symmetric layers and considering the loaded-vanes, the dispersion equation, interaction impedance and attenuation constant of helical slow wave structure with arbitrary tape width,supported by a radially stratified dielectric layer which is adjacent to the loaded vanes and enclosed by a metallic shell are obtained. The calculating results show good agreement with the experiment results. The theory can provide a strong basis for design of the helical s1ow wave structure.
By expanding the surface currents on the helix tape in a series of chebyshev polynomials, simulating the helical thickness with vacuum layer, equaling the rods with multiple azimuthally symmetric layers and considering the loaded-vanes, the dispersion equation, interaction impedance and attenuation constant of helical slow wave structure with arbitrary tape width,supported by a radially stratified dielectric layer which is adjacent to the loaded vanes and enclosed by a metallic shell are obtained. The calculating results show good agreement with the experiment results. The theory can provide a strong basis for design of the helical s1ow wave structure.
2003, 32(4): 409-412.
Abstract:
As one of the main properties of miniaturized HTSC filters, the complex couplings of HTSC Filter is proposed, and the analysis method of these complex couplings is studied. As the method, full wave technology is used to design the filter, and the couplings between or in resonators are studied by this method, then the effects to filtering response of these couplings are discussed. From the aspect of structures, the miniaturization procedure of HTSC Filter is under discussion with the complex couplings. There are some miniaturized structures and their simulated and tested results, and the fabricated filters are used in one real communication receiver.
As one of the main properties of miniaturized HTSC filters, the complex couplings of HTSC Filter is proposed, and the analysis method of these complex couplings is studied. As the method, full wave technology is used to design the filter, and the couplings between or in resonators are studied by this method, then the effects to filtering response of these couplings are discussed. From the aspect of structures, the miniaturization procedure of HTSC Filter is under discussion with the complex couplings. There are some miniaturized structures and their simulated and tested results, and the fabricated filters are used in one real communication receiver.
2003, 32(4): 413-416.
Abstract:
A new kind of NC system for machining shafts with non-circular cross-section is proposed in this paper. On the basis of a special micro-feed mechanism, which is driven by a simple cam, a new polar coordinate interpolation algorithm, which is based on affine transformation, is presented. By this method, shafts with any non-circular cross-section can be machined. Comparing with other similar systems, it possesses such advantages as:larger cutting force, better flexibility and higher precision et al.
A new kind of NC system for machining shafts with non-circular cross-section is proposed in this paper. On the basis of a special micro-feed mechanism, which is driven by a simple cam, a new polar coordinate interpolation algorithm, which is based on affine transformation, is presented. By this method, shafts with any non-circular cross-section can be machined. Comparing with other similar systems, it possesses such advantages as:larger cutting force, better flexibility and higher precision et al.
2003, 32(4): 417-419.
Abstract:
In this paper, according to knapback problem's characteristics, introduce zero-knowledge proof and unbiased coin cast protocols in theroy, and apply it to zero-knowledge proof and unbiased coin cast sucessfully by proving and analysing. More, we prove it's security theoretically and prove knapback has broad applicability by the extension of zero-knowledge proof, and so prove practicability of knapback problem with facts.
In this paper, according to knapback problem's characteristics, introduce zero-knowledge proof and unbiased coin cast protocols in theroy, and apply it to zero-knowledge proof and unbiased coin cast sucessfully by proving and analysing. More, we prove it's security theoretically and prove knapback has broad applicability by the extension of zero-knowledge proof, and so prove practicability of knapback problem with facts.
2003, 32(4): 420-424.
Abstract:
Real-time CORBA specification extends the scope of CORBA application to real-time field. A real-time CORBA prototype called rtORB is introduced. Then a test platform is proposed to evaluate real-time performance of rtORB. The delay and delay jitter are selected as test parameters to compare rtORB with other real-time CORBA product. The test results show that rtORB can support real-time application. A series of experiments on various operating systems are conducted to evaluate the influence that operating system and communication system impact on real-time CORBA. The test results are discussed in details and some suggestions directing how to build real-time application based on real-time CORBA are given.
Real-time CORBA specification extends the scope of CORBA application to real-time field. A real-time CORBA prototype called rtORB is introduced. Then a test platform is proposed to evaluate real-time performance of rtORB. The delay and delay jitter are selected as test parameters to compare rtORB with other real-time CORBA product. The test results show that rtORB can support real-time application. A series of experiments on various operating systems are conducted to evaluate the influence that operating system and communication system impact on real-time CORBA. The test results are discussed in details and some suggestions directing how to build real-time application based on real-time CORBA are given.
2003, 32(4): 425-429.
Abstract:
Based on the design of PMC-Sirra 7364 card device driver under Solaris 8 of UNIX system, the development technology of PCI device driver under UNIX system is discussed. It presents the idea of design, basic programming procedure, program structural framework, some useful functions and the system management commands. This method can be further applied to the development of CompactPCI device driver under UNIX, which provides a new approach to replacing PC industrial control computer with UNIX host to implement telecommunication and control devices with high stability and reliability.
Based on the design of PMC-Sirra 7364 card device driver under Solaris 8 of UNIX system, the development technology of PCI device driver under UNIX system is discussed. It presents the idea of design, basic programming procedure, program structural framework, some useful functions and the system management commands. This method can be further applied to the development of CompactPCI device driver under UNIX, which provides a new approach to replacing PC industrial control computer with UNIX host to implement telecommunication and control devices with high stability and reliability.
2003, 32(4): 430-432,450.
Abstract:
This paper briefly talks about the router's architecture and the method of switching data and puts forward a comprehensive improving scheme of queue grouping, feedforward monitoring and congestion controlling. Emulation shows that the adoption of the scheme is of certain practical value to improve the real-time property of network data switching a nd raise the throughput capacity. It also effectively keeps the phenomena of spreading of local port block to the whole network within limits.
This paper briefly talks about the router's architecture and the method of switching data and puts forward a comprehensive improving scheme of queue grouping, feedforward monitoring and congestion controlling. Emulation shows that the adoption of the scheme is of certain practical value to improve the real-time property of network data switching a nd raise the throughput capacity. It also effectively keeps the phenomena of spreading of local port block to the whole network within limits.
2003, 32(4): 433-436.
Abstract:
In this paper, the development trend and the general significance of the supervisory control system is analyzed, and also a design project of water supply's supervisory control and data acquistion system based on Web is introduced. The Web database adopts ASP technology to realize, and the long distance intelligent terminal uses MCU system. The user can supervise and control the water supply's equipments though the browser. The design has met with success in the system of railway water supply's supervisory control. Though the combination between control network and Internet, the corporation can achieve its automation with c ontrol, management, information and network together.
In this paper, the development trend and the general significance of the supervisory control system is analyzed, and also a design project of water supply's supervisory control and data acquistion system based on Web is introduced. The Web database adopts ASP technology to realize, and the long distance intelligent terminal uses MCU system. The user can supervise and control the water supply's equipments though the browser. The design has met with success in the system of railway water supply's supervisory control. Though the combination between control network and Internet, the corporation can achieve its automation with c ontrol, management, information and network together.
2003, 32(4): 437-439.
Abstract:
This papaer advanced that the client/server model of process is the bes t model for heterogeneous PLC communication service software on industrial Ether net in accordance with its special characteristics, and proposed an ideal of se amless connection between heterogeneous PLC networks in control systems by invok ing API function based on the model. The key technique used in communication ser ve software were analyzed in detail. Through the analysis of the program, the scienti fic practicability of the method is verified.
This papaer advanced that the client/server model of process is the bes t model for heterogeneous PLC communication service software on industrial Ether net in accordance with its special characteristics, and proposed an ideal of se amless connection between heterogeneous PLC networks in control systems by invok ing API function based on the model. The key technique used in communication ser ve software were analyzed in detail. Through the analysis of the program, the scienti fic practicability of the method is verified.
2003, 32(4): 440-443.
Abstract:
This paper discusses the problem in the credit management system of the universities and presents a security solution for them. The solution is based on the public key infrastructure technology. In the solution, the validation of user identity adopts the figure certificate, and the key data is figure validated. The data t ransmission adopts the secure socket layer. The authenticity, integrality and the undeniableness is realized. The credit management is the trend of the universities of China, and more and more universities put the credit system in practice. The implementation of the security solution, which assures the security of the credit management system, is of significant practical value.
This paper discusses the problem in the credit management system of the universities and presents a security solution for them. The solution is based on the public key infrastructure technology. In the solution, the validation of user identity adopts the figure certificate, and the key data is figure validated. The data t ransmission adopts the secure socket layer. The authenticity, integrality and the undeniableness is realized. The credit management is the trend of the universities of China, and more and more universities put the credit system in practice. The implementation of the security solution, which assures the security of the credit management system, is of significant practical value.
2003, 32(4): 444-446.
Abstract:
In this paper, according to the concepts of block diagonal dominance and generalize diagonal dominance, a new simple criterion for nonsingular block generalize diagonal dominant matrices (i.e. block H-matrices) are obtained by making use of the norm of submatrices of a given matrix only. Advantages are illustrated by numerical examples. And a necessary condition for nonsingular block Hmatrices is also presented. These results generalize the corresponding results in this area, and make a nodest contribution to this area.
In this paper, according to the concepts of block diagonal dominance and generalize diagonal dominance, a new simple criterion for nonsingular block generalize diagonal dominant matrices (i.e. block H-matrices) are obtained by making use of the norm of submatrices of a given matrix only. Advantages are illustrated by numerical examples. And a necessary condition for nonsingular block Hmatrices is also presented. These results generalize the corresponding results in this area, and make a nodest contribution to this area.
2003, 32(4): 447-450.
Abstract:
This paper considers a cold standby repairable system consisting of two dissimilar units and two repairmen. In the system, the life distributions of the two units are assumed to be the general distribution, but the repair distribution to be the exponential. By using the non-markov renewal process and the tool of the Laplace transform, we obtain some reliability quantities of the system,such as the distribution of the first failure, the availability and the frequency of failure of the system, etc.
This paper considers a cold standby repairable system consisting of two dissimilar units and two repairmen. In the system, the life distributions of the two units are assumed to be the general distribution, but the repair distribution to be the exponential. By using the non-markov renewal process and the tool of the Laplace transform, we obtain some reliability quantities of the system,such as the distribution of the first failure, the availability and the frequency of failure of the system, etc.
2003, 32(4): 451-454.
Abstract:
In this paper, a n-unit series repairable system with delay repair is studied,assuming that the systematic unit's life has exponential distribution, the dalay repair time and the repair time both have general distribution,some primary reliabilities indexes are obtained by using renewal process theory, such as reliability R(t), availability A, the mean time to first failure MTTFF and the mean times to failure of the system during (0,t],etc.
In this paper, a n-unit series repairable system with delay repair is studied,assuming that the systematic unit's life has exponential distribution, the dalay repair time and the repair time both have general distribution,some primary reliabilities indexes are obtained by using renewal process theory, such as reliability R(t), availability A, the mean time to first failure MTTFF and the mean times to failure of the system during (0,t],etc.
2003, 32(4): 455-457,472.
Abstract:
We have analyzed comparatively the optimum pricing of dividing up demand interval equally and pricing of fixed discount ratio on the aspects of division of interval and the consumer's surplus which be gained by a monopoly enterprise. The conclusions are as the following:the demand interval will be divided up by this means that a front interval is longer than a hind interval (the first interval is exceptional) if we price under fixed discount ratio; the two pricing method will both make a monopoly enterprise gain equal consumer's surplus on the whole.
We have analyzed comparatively the optimum pricing of dividing up demand interval equally and pricing of fixed discount ratio on the aspects of division of interval and the consumer's surplus which be gained by a monopoly enterprise. The conclusions are as the following:the demand interval will be divided up by this means that a front interval is longer than a hind interval (the first interval is exceptional) if we price under fixed discount ratio; the two pricing method will both make a monopoly enterprise gain equal consumer's surplus on the whole.
2003, 32(4): 458-463.
Abstract:
Disassembling the organizing process of enterprises alliance and enterprises group into four stages and establishing mathematical model of the organizing costs. The model of organizing cost for enterprises alliance and enterprises group in crisp case are given first in terms of references, then extending the model of organizing cost to the fuzzy sense suitable for practical case, which is fuzzified model's coefficients of terms related to time, the solution is attained by using fuzzy optimization technique and nonlinear dynamic programming to allocate time distribution of different organizing stages. The proposed method provides effective theoretic reference for enterprises to choose development strategies according to different practical situations.
Disassembling the organizing process of enterprises alliance and enterprises group into four stages and establishing mathematical model of the organizing costs. The model of organizing cost for enterprises alliance and enterprises group in crisp case are given first in terms of references, then extending the model of organizing cost to the fuzzy sense suitable for practical case, which is fuzzified model's coefficients of terms related to time, the solution is attained by using fuzzy optimization technique and nonlinear dynamic programming to allocate time distribution of different organizing stages. The proposed method provides effective theoretic reference for enterprises to choose development strategies according to different practical situations.
2003, 32(4): 464-468.
Abstract:
The paper constructs index system of enterprise core competence and applies fuzzy numbers as expression mode, combining with enterprises' practice, and put forward a fuzzy multi-attribute method of evaluating core competence on the basis of traditional evaluating ways. Meanwhile, a feasible mathematic model is founded, and through the instance analysis, showing the effectiveness, common-ability and scientific of this model.
The paper constructs index system of enterprise core competence and applies fuzzy numbers as expression mode, combining with enterprises' practice, and put forward a fuzzy multi-attribute method of evaluating core competence on the basis of traditional evaluating ways. Meanwhile, a feasible mathematic model is founded, and through the instance analysis, showing the effectiveness, common-ability and scientific of this model.
2003, 32(4): 469-472.
Abstract:
A method of stock price prediction is presented by hypothesis of stock market being non-linear dynamic system and analyzing method of chaos theory for chaos time series in this paper. Meanwhile, structures of radial basic function (RBF) network and pairs of training samples are determined by embedding dimension and delay time of reconstruct phase space respectively. Predicting results for real world stock time series show that the method is able to do effectively short-term prediction. In comparison with traditional forward feedback neural network (BP), the method can make better predicting performance, thus it can be widely used in stock price prediction.
A method of stock price prediction is presented by hypothesis of stock market being non-linear dynamic system and analyzing method of chaos theory for chaos time series in this paper. Meanwhile, structures of radial basic function (RBF) network and pairs of training samples are determined by embedding dimension and delay time of reconstruct phase space respectively. Predicting results for real world stock time series show that the method is able to do effectively short-term prediction. In comparison with traditional forward feedback neural network (BP), the method can make better predicting performance, thus it can be widely used in stock price prediction.
