Cognitive users can opportunistically utilize the unused spectrum resources without causing interference to primary user or affecting primary users communication. How to apply transmit beamforming to eliminate the primary user interference from cognitive user is the key. Firstly, the system model that the cognitive user network coexists with primary user network was developed. Then, an optimal beamforming algorithm that can completely eliminate the primary user interference from cognitive user was proposed. Compared with calssical zeroforcing beamforming algorithm, the proposed algorithm not only can eliminate the the primary user interference from cognitive user, but can improve the performance of itself. Finally, a performance simulation of the optimal beamforming algorithm was present. The result shows that the proposed algorithm has a better performance than zeroforcing beamforming.

The system of PDS(Pattern time Delay Shift) coding  communication can resist the multi\|path interfere well，but it can not resist noise in the channel well. Integrating TCM (Trellis\|Coded Modulation) code and PDS scheme can solve this problem by coding the source information with convolutional code and decoding the received sequence by soft\|decision Viterbi algorithm. Experimental results show that the proposed scheme gets about 3 dB coding gain, has better performance over the original decoding scheme.

An intelligent prediction model based on neural network and event sample database is proposed to solve the changes of traffic trends and the model failure after the introduction of new technologies and new business and decrease of prediction accuracy. The method has self\|learning function and continuing adaptive capacity on changes of traffic trends, events influence and the occurrence of new events,which can automatically adjust predictive parameters and update the event samples according to predictive error. Simulation results show that the method effectively improves the accuracy of traffic prediction, also has a high value for investment planning of the company's customer service center.

Fundamental theory judging by electromagnetic compatible, electromagnetic susceptibility and ego specifically for communication measures an instrument defend condition, new technique applying EMC (ElectroMagnetic Compatibility ) to develop in process in realization, Take that the basic testing demands and the electromagnetism environment of the Analog working as the main body,Measure having studied system, system inner shield. Structure, Shield way, component keep apart, problem such as, PCB（Printed Circuit Board）component preparation by screening proposes that systematically to physics in equipment to resolve a scheme. To communication the instrument designs that the antiinterference ability of element has had very big improvement effects basically.

In order to solve the difficult problem in car crash experiment to measure and record data, a dynamic image from a car crash detection, target identification and tracking algorithm is proposed. The algorithm uses binary images related to the target and background separation in the region of interest identified targets adjacent frames after the extraction of texture coordinates and feature matching; gives quantitative values of the matching strategy of integration of the weighted average by the criterion of minimizing error rate, etc. fusing the best weights; in the integration of quantitative value is defined as the basis of similarity, by decision threshold between the objectives of the adjacent frames to identify the greatest similarity combinations; which further raised with the same structure mapping principle to determine the best matching target adjacent frames. Experimental results show that the algorithm matching adjacent frames target accuracy than the traditional 5% increase in single\|feature methods, to more effectively track the target.

To implement the simple and portable characteristic for electrocardiogram machine,this design adopts the dry battery, which can provide not more than 3 V. The whole circuit is composed of amplifier AD623,TLV2254 and some resistance or capacitance. High precision enlarged for faint electrocardiogram signal,the fast elimination for polarized voltage and other functions are achieved by this circuit. It ensures the accuracy and steady of electrocardiogram signal,providing good situation for the flowing data sampling and analyzing for electrocardiograph(ECG) machine.In the same time, since using single power supply, it make power supply mode of electrocardiogram machine more simplify, make circuit more simple, promote the electrocardiogram machine further to portable direction.

In order to improve PSOs performance on complex engineering problems, it presents a variant of PSO（Particle Swarm Optimization)  called PSOPC (Particle Swarm Optimizer based on Predator-prey Coevolution), which takes into account the predatorprey behavior therefore high species diversity can be maintained as the whole population evolves. PSOPC is then applied to schedule RFID（Radio Frequency Identification） networks by assigning time slot to RFID readers online whereby reader collision can be minimized to ensure the current operation of the RFID system. Simulation results demonstrate that the PSOPC algorithm is more feasible and efficient than PSO in solving this realworld problem.

Well-developed, concise and accurate software project schedule planning is the key to successful software development.The current schedule planning mainly uses manual methods, draw up slow, lack of scientific characteristics and so on, the article bases on the network chart, the use of work breakdown structures, program evaluation and review technique, critical path method, gantt chart and other technical studies. It illustrates preparatory work firstly, including the definition of activities, activity sequencing, drawing the network chart, gives time parameters derived formula, including estimation the activity time, estimation the project cycle, calculation activity time, and describes preparation process by examples. The results show that the method of software project scheduling planning has high practical value.

The complexity of many\|light global illumination has been limiting its use only to off\|line applications.We firstly approximates many light by a single brighter light,and the light paths are traced from this representative light, and VPLs(Virtual Point Lights) are generate at the vertices of these paths. Then, those VPLs are used to illuminating the scene, thus single\|bounce indirect illumination can be rendered on currently available graphics hardware at real time rates. The direct illumination of other lights are pre\|computed and stored in lightmaps, the contribution of dynamic lights and the resulting indirect illumination from that are computed and then combined with the lightmaps. At last, with little sacrifices in quality, the rendering of global illumination of the many\|lights scene with dynamic representative light and dynamic camera achieves interactive rates.

Textures usually are generated by computer. However, how to control the texture evolution to generate a large amount of beautiful, vivid textures with computer has become a very interesting and challenging problem in the research on computer graphics. A new normative expression which unifys all kinds of Arnold transformation and its inverse transformation is proposed. To extract some pattern of representation national style from xinjiang database, by normative expressions of the forward and inverse transformation of the pattern matrix, the process of texture generation are described and obtain a great of different style pattern. Whether forward transformation or inverse transformation, Arnold transformation can be calculated as the inverse matrix usually calculated as positive and negative transformation matrix, which cannot be limited on (mod N). And a method is extended to generate texture patterns of national style. The method is simple and easy to generate a large amount of different style patterns rapidly. Based on the introduction of the evolution generation of national pattern, more design style patterns are laid the foundation.

The problem of parameterdependent robust stabization for stochastic timedelay systems with polytopic uncertainties is investigated due to the much more conservatism of the existing results. By using Lyapunov stochastic stability and constructing parameterdependent Lyapunov functions, parameterdependent robust stability sufficient condition is obtained in terms of LMIs(Linear Matrix Inequalities). The proposed method is suitable to engineering due to its simple calculation and accurate result. An example is given to illustrate the effectiveness of results.

 Edge information in remote sensing image provides a rich content for image understanding.An algorithm to calculate the fractal dimension of a single pixel by using the model of DFBR（Discrete Fractal Brown Random field)is designed.This algorithm is programmed by Matlab. Map the image gray level space to the fractal dimension of space, and then transform and edge detection in the latter one. Spatial distribution of surface features and the differences of image texture structure and edge features, make the selected window size used to calculate fractal dimension as the key. Select part of the high spatial resolution remote sensing image in the study area as the experimental data. By calculating the fractal dimension of pixels in different windows, the calculation window of the best edge detection results is required. The results show that this algorithm is more suitable for remote sensing data. And the accuracy of his remote sensing image extraction is improved.

Aim at the problem of lacking portfolio and integrate about the current monitor and alarm systems, an embedded intelligent monitor system based on ARM（Advanced RISC Machines） was proposed. The system uses an embedded WinCE operating system,S3C2440 processor of ARM and 13 million pixels CMOS（Complenmentary Metal-Oxide-Semiconductor） camera, which are composed of video capture device. The combination of the background subtraction and frame difference method,the method of adaptive updating background was proposed to detect moving object. Finally, the alarm information is sent by way of mobile email and mobile phone connected through SMTP（Simple Mail Trasport Protocol） and MIME（Mutipurpose Internet Mail Extension） transmission protocol. When someone broke into the monitoring sites,the systems can alarm on their own, and transmits the captured video information to the user hand side through the network. The experimental results showed that the intelligent monitor system has the high efficiency, rich interfaces and the higher security and reliability.

In order to solve the problem of poor effect in image stitching caused by the rotation, and the differences in light intensity, an image stitching algorithm based on the feature points is proposed in this paper. The algorithm first determines the overlapped area in images using the method of phase correlation, then extracts the corners with improved Harris detecting algorithm,and the feature point matching pairs acquired by NCC（normalized cross correlation） method, finally we are able to complement the fusion of stitching images by using the step-in-step-out method. The experiment results show that the algorithm improves the speed and accuracy of image stitching effectively.

 In order to prevent network fraud and identity theft and to increase authentication security,  an intelligent authentication method based on neural network and multi\|factor was improved. Based on analysis of the composing and distinguishing of identification, this method focuses the customers action.An application model which can be used in e\|commerce system is designed. Resolved the problem of traditional authentication method which depends on a few of identity factors， identity theft can be prevented, security of authentication is increased.