The large-scale applications of M2M(Machine to Machine) service will have great impacts on the current network in QoS，IBP(Interrupt Bernoulli Process) is used to model M2M services with batch arrival,then a discrete time queue model-IBP/Geom/1/K is established and solved.Different from typical IBP model,the number of arrival packets each time is a batch,instead of one.Mathematical models with different burstiness are used to describe the batch,then we obtain the steady-state probability of queue length in the probability domain,furthermore the throughout and packet loss ratio and other performance metrics are derived.System performance with single arrival and  with batch arrival under the same queuing intensity are compared.The results indicate that the system performances decrease as the burstiness of the number of arrival packets every batch increases and the queuing performances with batch arrival are worse than that with single arrival under the same queuing intensity.For M2M small data services with delay-tolerance characteristics,we can increase cache at the cost of increasing delay to improve the system throughput and decrease the blocking rate.

In order to compare the advantage and shortcoming  of two Doppler estimation algorithms for OFDM（Orthogonal Frequency-Division Multiplexing） in underwater acoustic communication,a lacustrine experiment was carried out.The replica correlation time delay difference estimation algorithm is simple and easy to achieve,and the null subcarriers algorithm is complex and accurate.The same experimental dates are estimated and compensated based on the two algorithms,and error bit rates are compared.The results show that the two methods have mainly uniform error bit rates in slowvarying UWA(Underwater Acousitic) channels,but the null subcarriers algorithm is more stable in fast time-varying channels.

In order to obtain the desired signal of dualaxis inclinometer from the complex automotive environment，a LMS（Least Mean Square） algorithm for reducing the convergence speed to process the related data is presented,it presents a self-adaptation step LMS algorithm by establishing the non-linear relationship between the step and the error,so as to the step is interrelated with the true signal for reducing the algorithm sensitivity to noise.To cope with the problem of superiority and correctness for applying to the vehicle vibration signal extraction,the learning curves of algorithm are gained by inputting the data with additive white noise of time domain.A comparison of them is made by convergence speed and computation complexity index.The result shows that the novel algorithm is reliable．The vehicle acceleration signal extraction experiment shows that it  removes the vehicle environmental impact of the dual-axis inclinometer,and  restrains its own output noise.
Based on the self-adaptive step LMS algorithm,it proposes and implements a scalable FIR （Finite Impulse Response）filter structure on FPGA(Field Programmabl
e Gata Array) .By reusing of fist-order filter unit to complement the multi-stage cumulative,it achieves 256 filter experiments on the largest FPGA chip.The result indicates that the FIR filter structure full use of chip resources,processing speed highly,reliability highly and apply for the low speed for vehicles attitude measurement system.

In order to improve the TV-guided missile penetration ability and kill rate.The method of guidance of  TV image tracking is studied.TV-guided weapon system is based on image processing, automatic control and information science. It extracts the location information of missile image from TV image, and guides the missile to attack the target. TV-guided weapon system can divide into two big class that is TV-command guided system and TV-target seeking  guided system. According to the method of pick up target position information from target image， the pattern of image track can divide into: the contrast track pattern and the image interrelated pattern. Through the study of several track methods in common usage currently: the edge track,the heart of rectangle track,the peak value track and the absolute bad method of related track, it has the positive guidance to improve the TV-guided missile penetration ability and kill rate.

When the traditional algorithms are applied to estimate DOA（Direction of Arrival） of signals with character of cyclostationarit using MIMO（Multiple Input and Multiple Output） radar system,the accuracy is very disappointed and even the algorithms will fail.So cyclic-music algorithm is used to estimate DOA in MIMO radar system.The theory of cyclic statistics is used to calculate the cyclic correlation function between the received signals,to construct correlation matrix of the tectonic cycle,and decompose the singular value.After searching the spectrum peaks,we can obtain the azimuth information.The theoretical deduction and simulation experiments show that for the signals with the characteristics of cyclostationarity,MIMO radar is effective and gets better accuracy by using Cyclic-Music algorithm.

In order to solve the problem in choosing wining structure of building the FTTH (Fiber To The Home)network under the uncertain conditions of access users,we use qualitative and quantitative analysis methods to compare  cross-connecting distribution line type of ODN(Optical Distribution Network) structure and direct connecting distribution line type of ODN structure,find out that  it has a stable access ability and a wide covering ability and a flexible expansion ability,so that is appropriate for the application in the competitive telecommunication market，has certain reference significance during large-scale construction projects on the FTTH network.

In order to realize property rights of welding robot welding function, a new type of robot controller named KeMotion robot controller is described in this article. And based on  the welding integrated application a controller platform is realized, including the control bus structure, the hardware configuration and the application command mode of the welding system.And on the platform the development of robot welding function is realized, experiment verified the validity of the welding function and  the  requirements of the welding process.

In order to ensure the measuring accuracy of resistance type sensor,a gas detecting system based on constant current circuit has been designed.The system uses constant current to make the gas sensor’s resistance  linear  with output voltage,so that the detecting  resistance accuracy  of gas sensor is improved.The  constant current circuit consists of reference voltage supply,current negative feedback circuit,voltage stabilizing diode,and load resistance.After calibration,its measuring range is 0～1 000 kΩ,andcorresponding average measuring error is less than 0.027 2%.By using SnO₂ sensor to detect 0～2 05357  mg/m³ ethanol gas,the calibration results show that the resistance decreases exponentially as the concentration increases,and the maximum detection error is less than ±1.5%.

For nearfield wireless energy transmission system,if the input signal’s waveform and the average power are changed,it may have some impact on the transmission efficiency of the signal energy.But  no similar researches published at home or abroad.In order to fill the gaps in the research areas,in combination with theoretical analysis and experimental verification,we studied  the relationship between signal’s waveform and power with its energy on wireless energy transfer efficiency.We respectively select the square wave,sine wave and triangular wave as the input waveform.Through horizontal and vertical analysis,we come to  two conclusions:Same input waveform,same energy transfer efficiency.The efficiency will not change if the input power is changed; Different input waveform,different efficiency.The wireless energy transfer efficiency ratio  is:square wave/sine wave/triangular wave=2.622∶1.768∶1.

To solve the positioning problems between two moving bodies,we  presents a kind of positioning method based on phase change rate in two-dimensional model,conducts and gives the results of observability analysis.Several classical nonlinear filtering algorithms are introduced and the  EKF（Extended Kalman Filter）、UKF（Unscented Kalman Filter）、PF(Particle Filter） are applied to the Location model.Simulation results show  that UKF takes the least time and PF is the
 algorithm with best accuracy.

In order to transmit network datas safely among terminal devices,the design of secure file transfer system based on embedded Linux is introduced.The architecture of secure file transfer system is  introduced,the way of designing system software of server and client is explained based on SSL（Secure Socket Layer）protocol,and
system security is analyzed.The file between server and client is encrypted through the ssl protocol during the transportation,and the server is connected by several clients because of multithreading.The design can help solve the problems in transmitting network datas safely and facilitate to develop network security in the mobile internet.

To increase the efficiency of learning,we describe the instructional application of virtual experiment by focusing on two types of learning systems which are knowledge-based learning system and ability-based training system.Feedbacks of learners,such as experimental operation accuracy and time intervals are mapped to flow theory.At last,we suggest self-adaptive adjustment strategies toward each learning systems which learners perceive significantly higher level of learning satisfaction.


In order to study the competitive interactive behavior of the payment institutions,we review the influence of user heterogeneity for payment service innovation strategy based on the transfer cost.We also analyze the formation conditions of the local progressive evolutionary stable strategy through evolutionary game，and prove that epayment market Innovative behavior will be intensified or weakened when there is an evolution equilibrium pathway.We offer a theoretical framework for researching innovation competition interactive between payment institutions.

ELM（Extreme Learning Machines） always works inefficiently on large-scale and high-dimension datasets for huge memory and computation costs.We prepose a novel parallel algorithm for training ELM quickly based on “divide and conquer” strategy.It dispatches large-scale datasets to a cluster of computing nodes by utilizing special binary cascade architecture,and then updates weights of SLFN(Single Hidden Layer Feed-forward Neural Network) in parallel.Theoretical analysis proves that the new algorithm converges to the best least-square solution monotonously with finite steps.Preliminary experimental results show that the new algorithm  has good generalization ability, excellent speedup ratio and parallel efficiency.

In order to solve the separation between personal communication key and group communication key in traditional Ad Hoc network,we propose a encryption scheme in Ad Hoc network which is based on identity-based cryptosystem,and it has ability of clusterlevel and revocation mechanism.In the scheme,the sender only use one key to specify the access permissions of the encrypted data by using the cluster level,and it can revoke any subset of the members to complete the personal communications or designated members of the group communication.

To solve the problem of dual aircrafts path planning for forest fire rescue in complex environment,artificial immune algorithm was proposed for three dimensional flight path planning.Referring to the method of using artificial immune algorithm to plan robot path,three-dimensional flight paths for dual aircrafts were designed by considering the special characteristics of plan and the requirements of dual aircraft path planning in the consideration of taking off at different sites and arrival at desition at the same time,and main factors of the algorithm were analyzed and estimated for finding the optimal parameters of planning track for designing safe and efficient flight path.Results show that the method is capable of planning a dual aricrafts flight path in complex environment.Artificial immune algorithm with optimal parameter will  be able to quickly and efficiently planning three-dimensional tracks and  enrich the route planning method.

In order to reduce the intensity of work for the person of computer laboratory and improve the efficiency of laboratory maintenance,combining with the technology of PXE(Preboot Execute Environment) network root,which is proposed that a program for computer laboratory maintenance is based on net root.The method for different network adapters rooting,different types of computer system integrate into a single root file is otudied,forming the strategy of computer laboratory maintenance which adept different environment.After the application of various computer labs,confirming the efficiency and practical of this maintenance strategy so as to provide a reference with management for computer laboratoty.