针对《现代交换技术》课程实验教学,设计了一种模块化实验教学方案.该方案是将交换系统分解成功能相对独立的模块,通过对不同模块的讲解、操作,让学生掌握相应的理论知识及应用.实践教学表明:该教学设计方案能有效地提高学生对相关知识的整体理解和掌握.

传统激光遥感图像边缘锐化方法,在图像边缘处理时会增加局部噪声、尤其在光照不均匀条件下,容易导致图像边缘失真.针对传统方法的不足,提出一种基于模糊理论的海量激光遥感图像边缘锐化方法研究,首先计算激光遥感图像边缘的锐化阈值,依据模糊特征分别建立边缘和噪声的隶属度函数,求解激光遥感图像像素相对于边缘和噪声的隶属度,并完成对图像边缘锐化系数的调整和控制,最终实现光照不均匀条件下,对海量激光遥感图像边缘的锐化处理.仿真实验结果证明,提出方法能够有效提高激光遥感图像的空间分辨率,滤除图像边缘的干扰点,在保护图像细节特征的同时,增加了图像边缘的清晰度.

针对目前扫频信号源的扫速调整复杂性问题,提出了一种可快速调整步进长度的方法,以Alter公司CycloneⅡFPGA为核心完成DDS直接数字频率合成计,以ATmega128 AVR单片机为控制器调整频率控制字K值,从而改变DDS累加器的相位增量,完成步进单位和输出频率的设置.测试表明,本系统不但能以10nHz(n=0,1,2,3,4,5,6)为步进单位从1Hz～11MHz循环输出,也可设置1Hz～11MHz间任意整数的点频输出,而且大大降低了误差,输出信号波形质量优良,较传统扫频信号源能更好满足不同精度的频率特性测试需求.

针对单列像素结构的压缩成像系统中存在过度采样和目标移动过程中可能暂停的问题,提出了一种改进的基于线性传感器的移动目标压缩采样方法.该方法通过较高的采样频率不断地获取压缩测量值,当压缩测量值样本数达到一定数量后,通过比较的方式来判断后续测量值是否有效;如果测量值保持不变,此时表示目标处于暂停状态,判别测量值无效并丢弃;反之,保留并继续压缩采样.理论分析与实验仿真结果表明,采用本文方法对移动目标进行压缩采样时,能够有效的消除目标暂停而导致的过度采样问题;并利用传统的重构方法能够有效地恢复出目标完整的图像.

Integrated operational amplifier is an important device in the analog electronic technology course. Solving the output voltage of integrated operational amplifiers when the input voltage and values of devices are known is a key knowledge point. The output voltages of integrated operational amplifiers are solved using the combination of virtual short circuit and virtual open circuit, and theoretical derivation method. The summery of the two methods are given to help students master them well.

针对WSN及轮廓检测系统的特点,给出了一种基于WSN的数据处理方法.该方法首先通过主分量分析提取样本特征,然后采用叠加方式对不同样本的特征进行融合,给出数学模型;并以此模型为基础,提出一种新的基于稀疏表示的多类融合样本中特定目标识别算法,该算法根据超完备字典下主要非零系数的分布情况识别出特定目标;数值仿真与实验结果验证了本文算法的有效性,综合性能优于传统方法.

针对应用型高校电子信息类专业的人才培养要求,结合Protel课程特点,探究了基于案例教学法的Protel课程教学改革.教学实践表明:该教学模式不仅激发了学生的学习积极性,而且提高了实际动手能力和工程素养,取得了良好的教学效果,实现了课程教学目标.

针对物体外形轮廓特征,给出一种基于主动式光电传感器的轮廓特征提取方法,该方法是通过物体经过传感器组视场时各传感器输出状态来获取特征信息;并结合稀疏表示理论,提出一种基于稀疏表示的目标轮廓识别算法,该算法是将测试样本投影到稀疏域,依据系数的稀疏性对测试样本进行分类.数值仿真与实验结果表明:采用光电传感器组获取的信息能够有效地描述物体;且通过本文算法对不同类型物体进行分类,具有良好的识别效果;对不同条件下获得的特征样本进行分类时,具有良好的鲁棒性.

本文针对《程控交换原理》课程,重点分析了该课程实践教学过程中主要存在的一些问题和不足,给出了对该课程实践教学进行改革的一些新的思路和方向。

According to the output signal characteristics of the profile detecting system,special object recognition method based on sparse representation,combined with the theory of sparse representation and principal component analysis,is proposed.First,using principal component analysis,the method extracts the main components of the sample signal in order to eliminate redundant information.Second,the signal is transformed into the same size of the feature vectors,which is then projected to the lower dimensional space to construct a dictionary.Finally,the testing samples are sparsely represented and recognized by the dictionary.Numerical simulations and experiments show that the proposed method has good classification effect in lower dimensional space,and good robustness for the system with some damage sensors in the actual situation.

针对目前通信原理实验教学不够灵活等不足问题,提出了一种新的实验教学方法,该方法主要通过前期仿真,然后学生自己搭建测试系统的形式,来完成实验内容。教学实践表明:该教学方法不仅能够很好地完成实验教学内容、加深学生对所学内容的掌握,同时也大大提高了学生的动手能力、激发了学生实验兴趣和理论学习积极性。

It's indicated that the result of Nash equilibrium of noncooperative power control game in multicell wireless data networks can't be Pareto optimality.A new power control algorithm based on best equal SIR is presented for multicell wireless data networks.Numerical results suggest that this algorithm can improve capability of the system,make terminals enjoy higher utility and lower transmit power,and the use of wireless resources become more rational and impartial.

In view of the traditional electric power and electronic technology experiment of deficiency, combined with DC/DC test examples to explain the switching power supply modules in parallel experiments the specific knowledge points, each knowledge point of the circuit design, the system functions, the specific test procedures and test methods. And given the experiment software design flow chart and measured data.

In view of the deficiency of traditional motor drive experiment,taking a DC shunt motor speed control experiment as an example,this paper describes the PWM control of DC motor speed control system design of experiments and specific knowledge points,each knowledge point circuit design,system functions,the specific experimental procedures and experimental methods,and gives the experimental procedure flow chart.

In order to avoid the disadvantages of low complexity and single generating scheme of the current chaotic spread spectrum sequence,a efficient scheme of generating chaotic spread spectrum sequences based on embedded model is put forward in the paper,and the generating chaotic sequences will be applied to the direct sequence spread spectrum system.The theoretical analysis and simulation results show that the scheme of generating chaotic sequences are not only high complexity,good secrecy,good performance,and can also produce large amounts of chaotic spread spectrum sequence,greatly improves the efficiency and provide the potential for applying.

In order to avoid the deficiency of the conventional schemes for code acquisition,this paper proposes a improved scheme for code acquisition,which could remove the influence of modulation information and random data by cycle-and-add property. At the same time,in order to improve signal to noise ratio at the receiver,the method of feedback circuit is adopted to suppress noise influence. The theoretical analysis and simulation results show that the proposed scheme could effectively improve detection probability and false alarm probability of code acquisition system,and thus enhance reliability of the system.

In order to avoid the influence of jamming signal on the performance of acquisition system,this paper proposes a new scheme of code acquisition,which is based on singular value decomposition that decomposes and reconstructs the received signal to suppress jamming,after this process,the signal enters acquisition system.The theoretical analysis and simulation results show that the method is capable of suppressing jamming,improving acquisition performance and stability of system.

An algorithm for ECG signal detection and classification based on wavelet aided M-classification SVM is proposed by this paper.The output of Wavelet transform of ECG signal is used as the input of M-classification SVM because of wavelet’s excellent features in time-frequency domain.The results of experiments show that it could recognize ECG signals effectively on the combination of wavelet and M-classification SVM.

In order to remove the bias,a noise ICA algorithm based the noise model is introduced and the multi-user detection method based this algorithm is deduced. The algorithms are simulated to compare the performance of the two kinds of multi-user detection methods which are based the normal ICA algorithm and the noise ICA algorithm. Simulation results indicate that noise ICA algorithm method has better performance.

In the spread spectrum communication systems,the conventional scheme of slide-correlation acquisition has a disadvantage that it needs a long time for the systems to acquire the synchronization with the received PN code sequence,especially when the sequence length is longer.In order to solve this problem,this paper proposes an improved scheme of code acquisition,which achieves PN code acquisition by three parallel circuits,the improvement is that the scheme isn't confined by the code length,but also effectively reduces false alarm probability,and the circuit is simple.The theoretical analysis and simulation results show the improved scheme shortens systemic acquisition time.