基于散射函数的一种微波非线性电路建模新方法
|
摘要
本论文课题来源于“微波毫米波测试仪器基础研究”项目,作为其中的一个子课题,子课题题目为“微波非线性电路建模方法的研究”。
随着微波半导体电路在通信、雷达、电子和仪器仪表等方面的广泛应用,如何准确和高性能的设计和分析微波电路是非常关键的。半导体器件模型是影响电路设计精度的最主要因素,电路规模越大、指标和频段越高、对器件模型的要求也越高。因而准确的器件模型对提高射频和微波毫米波电路设计的成功率、缩短电路研制周期是非常重要的。根据微波电路的工作条件,它常常工作于高功率大信号条件下,电路状态处于强非线性情况,传统的小信号S参数等表征和建模方法已经不能准确表征其模型,而器件大信号模型不易获取。伏特拉级数、幂级数及等效电路方式常常表征电路的弱非线性状态,强非线性电路状态也不能准确表示。
论文在总结传统大信号电路的表征、测量方法和建模方法的基础上,分析了各种测量和建模方法的适用条件、特征和优缺点,并分析了对于微波电路建模的一般步骤。针对以往大信号模型的一些问题,引入了非线性散射函数这一表征微波大信号电路的频域黑箱模型。
论文研究了非线性散射函数模型的线性化,上下标标注方式的引入,各个非线性散射函数元素物理含义的分析和描述,散射函数与传统小信号S参数的相互推广。对其它非线性网络函数的定义,非线性散射函数与其它非线性网络函数的相互关系,网络连接的非线性网络函数和子网络的非线性网络函数的关系进行了研究和探讨,研究结果为这一模型的正确使用奠定了理论基础。
此外,开发了基于数字示波器的非线性散射函数提取系统及系统误差的校准方法,为准确提取功率器件的非线性散射函数模型提供了实验条件,并经过测试得到了功率场效应管和二极管的非线性散射函数变化趋势,测试结果正确反映出功率器件随激励信号功率增大的非线性输出变化规律。
研究了非线性散射函数模型与近年来美国Agilent公司提出的X参数模型从定义、性质、提取方法和使用等方面的联系和区别,为散射函数模型的研究和应用提供了有益的参考和比较。
在功率器件的非线性散射函数测量数据基础上,采用支持向量机这一新的机器学习方法对散射函数模型做预测建模,得到了较好的预测精度,分析了支持向量机形式选择、核函数选择、参数选择对建模精度的影响。
最后采用场效应管和肖特基势垒二极管的散射函数模型来设计微波功率放大器和倍频器,验证了模型的适用性。
与传统大信号模型相比,这一模型的优点主要在于易于与小信号S参数相兼容,通过这一模型可以建立起大信号和小信号工作条件下的微波电路的统一模型;与目前国外的非线性网络测量系统不同,测量方法采用数字示波器为基础的时域测量方法,测试简单方便,采用通用测试夹具可以对各种微波半导体电路用统一的方法来测量提取散射函数模型;这一模型是频域黑箱模型,不需要预知微波半导体器件和电路的电路结构就可以提取出散射函数模型,模型不止包含本征模型部分,对于寄生参数的影响也被包含在了这一黑箱模型中,并可以将其用于微波电路的设计中,方法通用,精确度高。
实践证明,非线性散射函数这一新的微波器件大信号模型,对有效提高微波非线性电路的设计效率、提高设计准确度和性能具有重要的意义。为微波大信号电路的高效设计和研发提供了一种新的思路。
此外,对于散射函数提取中相位一致性不好问题的改善;夹具去嵌问题的研究和实现;支持向量机对部分非线性散射函数建模精度不高以及如何将散射函数模型设计为可供通用CAD软件调用的统一微波电路大信号模型等问题,仍是需要深入研究和探讨的问题。
As a sub-topic named“Research of microwave nonlinear circuit modeling method”, this paper is derived from the project of“Basic research of microwave and millimeter wave test equipment”.
With the wide use in communication, radar, electronics and instrumentation,how to accurately design and high-performance analyze microwave semiconductor circuits is very important. Models of semiconductor devices are the main factor to affect the accuracy of the circuit design. Larger the circuit , higher the frequency, to the higher requirements of the device models. It is critical of the accurate device models to improve the design of RF and microwave millimeter-wave circuit and shorten the development cycle. According to the working conditions of microwave circuits, which are often working in high-power and large-signal, the circuit is in a strongly nonlinear. Traditional characterization and modeling methods of S parameters at small-signal can not represent them accurately, while the large-signal device model is not easy to obtain. Volterra series, power series and the equivalent circuit approach is often used in a weak non-linear state of the circuit, to the strong non-linear circuits it can not be an accurate representation.
Based on the traditional large-signal circuit characterization, measurement and modeling methods, it is analyzed that the application, characteristics, advantages and disadvantages of a variety of methods. General steps of microwave circuits modeling are summarized. As some problems of traditional modeling methods, the frequency domain black-box model of nonlinear scattering function is introduced.
First, the research is focused on the linearization of nonlinear scattering function model, introduction of the upper and lower standard labeling methods, analysis and description of the physical meaning to the element of nonlinear scattering function and the interchangeable of the traditional S parameters. Secondly, it is concluded that other description of the nonlinear network functions, the relationship between nonlinear scattering function and other nonlinear network function and the relationship between nonlinear network function of the network link and the sub-network. The results of them have laid a theoretical basis for the proper use of this model.
Then, it is developed the extraction system of nonlinear scattering function based on the digital oscilloscope and the calibration methods of system errors. This system provided the experiment conditions to the extraction of nonlinear scattering function model of power devices. Nonlinear scattering function of power FETs and diodes can be tested and the results accurately reflected the nonlinear variation with the increase of input power to the power devices.
Links and differences of the model of nonlinear scattering function and the model of parameters X are proposed by Agilent recently are analyzed. Through the comparison of the description、characteristic、extraction methods and the applications between them, it provides a useful reference and comparison to the study and application of nonlinear scattering function.
Based on the test data of nonlinear scattering function of power devices, the new machine learning method of support vector machine is used to predict the model of nonlinear scattering function and the good predict accuracy is got. Then, it is summarized that the affection to the modeling accuracy from the selection of the support vector machine, kernel function and other parameters.
Last, the model of nonlinear scattering function of FETs and Schottky barrier diodes are used to design the microwave power amplifier and frequency multiplier. It is verified that the applicability of the large-signal model.
Compare with the traditional large-signal model, the advantages of this black-box model can be generalized. The first is the compatibility with the parameters S at small signal. Through this, the unified microwave circuit model can be established between the working conditions of large-signal and small-signal. Different from the foreign NNMS(Nonlinear network measurement system), the second is the time domain measurement method based on the digital oscilloscope. This method is simple and convenient. Adopting the common test fixture, a unified method can be used to exact the nonlinear scattering function model. The last, the model is frequency-domain black-box model, it can be extracted which is not needed to know the structure of microwave semiconductor devices and circuits in advance. The model not only contains some of the intrinsic part, and the affection of parasitic parameters is included in this model. It can be used in the design of microwave circuits conveniently and accurately.
It can be proved that this new large-signal model of microwave devices can improve the design efficiency, improve design accuracy and performance effectively.
In addition ,some problems, such as how to improve the consistency problem to the phase of nonlinear scattering function; modeling accuracy is not high to the part of nonlinear scattering function using support vector machine and how to design a unified large signal CAD(Computer aided design) model with the model of nonlinear scattering function, still requires in-depth study and explore.
随着微波半导体电路在通信、雷达、电子和仪器仪表等方面的广泛应用,如何准确和高性能的设计和分析微波电路是非常关键的。半导体器件模型是影响电路设计精度的最主要因素,电路规模越大、指标和频段越高、对器件模型的要求也越高。因而准确的器件模型对提高射频和微波毫米波电路设计的成功率、缩短电路研制周期是非常重要的。根据微波电路的工作条件,它常常工作于高功率大信号条件下,电路状态处于强非线性情况,传统的小信号S参数等表征和建模方法已经不能准确表征其模型,而器件大信号模型不易获取。伏特拉级数、幂级数及等效电路方式常常表征电路的弱非线性状态,强非线性电路状态也不能准确表示。
论文在总结传统大信号电路的表征、测量方法和建模方法的基础上,分析了各种测量和建模方法的适用条件、特征和优缺点,并分析了对于微波电路建模的一般步骤。针对以往大信号模型的一些问题,引入了非线性散射函数这一表征微波大信号电路的频域黑箱模型。
论文研究了非线性散射函数模型的线性化,上下标标注方式的引入,各个非线性散射函数元素物理含义的分析和描述,散射函数与传统小信号S参数的相互推广。对其它非线性网络函数的定义,非线性散射函数与其它非线性网络函数的相互关系,网络连接的非线性网络函数和子网络的非线性网络函数的关系进行了研究和探讨,研究结果为这一模型的正确使用奠定了理论基础。
此外,开发了基于数字示波器的非线性散射函数提取系统及系统误差的校准方法,为准确提取功率器件的非线性散射函数模型提供了实验条件,并经过测试得到了功率场效应管和二极管的非线性散射函数变化趋势,测试结果正确反映出功率器件随激励信号功率增大的非线性输出变化规律。
研究了非线性散射函数模型与近年来美国Agilent公司提出的X参数模型从定义、性质、提取方法和使用等方面的联系和区别,为散射函数模型的研究和应用提供了有益的参考和比较。
在功率器件的非线性散射函数测量数据基础上,采用支持向量机这一新的机器学习方法对散射函数模型做预测建模,得到了较好的预测精度,分析了支持向量机形式选择、核函数选择、参数选择对建模精度的影响。
最后采用场效应管和肖特基势垒二极管的散射函数模型来设计微波功率放大器和倍频器,验证了模型的适用性。
与传统大信号模型相比,这一模型的优点主要在于易于与小信号S参数相兼容,通过这一模型可以建立起大信号和小信号工作条件下的微波电路的统一模型;与目前国外的非线性网络测量系统不同,测量方法采用数字示波器为基础的时域测量方法,测试简单方便,采用通用测试夹具可以对各种微波半导体电路用统一的方法来测量提取散射函数模型;这一模型是频域黑箱模型,不需要预知微波半导体器件和电路的电路结构就可以提取出散射函数模型,模型不止包含本征模型部分,对于寄生参数的影响也被包含在了这一黑箱模型中,并可以将其用于微波电路的设计中,方法通用,精确度高。
实践证明,非线性散射函数这一新的微波器件大信号模型,对有效提高微波非线性电路的设计效率、提高设计准确度和性能具有重要的意义。为微波大信号电路的高效设计和研发提供了一种新的思路。
此外,对于散射函数提取中相位一致性不好问题的改善;夹具去嵌问题的研究和实现;支持向量机对部分非线性散射函数建模精度不高以及如何将散射函数模型设计为可供通用CAD软件调用的统一微波电路大信号模型等问题,仍是需要深入研究和探讨的问题。
As a sub-topic named“Research of microwave nonlinear circuit modeling method”, this paper is derived from the project of“Basic research of microwave and millimeter wave test equipment”.
With the wide use in communication, radar, electronics and instrumentation,how to accurately design and high-performance analyze microwave semiconductor circuits is very important. Models of semiconductor devices are the main factor to affect the accuracy of the circuit design. Larger the circuit , higher the frequency, to the higher requirements of the device models. It is critical of the accurate device models to improve the design of RF and microwave millimeter-wave circuit and shorten the development cycle. According to the working conditions of microwave circuits, which are often working in high-power and large-signal, the circuit is in a strongly nonlinear. Traditional characterization and modeling methods of S parameters at small-signal can not represent them accurately, while the large-signal device model is not easy to obtain. Volterra series, power series and the equivalent circuit approach is often used in a weak non-linear state of the circuit, to the strong non-linear circuits it can not be an accurate representation.
Based on the traditional large-signal circuit characterization, measurement and modeling methods, it is analyzed that the application, characteristics, advantages and disadvantages of a variety of methods. General steps of microwave circuits modeling are summarized. As some problems of traditional modeling methods, the frequency domain black-box model of nonlinear scattering function is introduced.
First, the research is focused on the linearization of nonlinear scattering function model, introduction of the upper and lower standard labeling methods, analysis and description of the physical meaning to the element of nonlinear scattering function and the interchangeable of the traditional S parameters. Secondly, it is concluded that other description of the nonlinear network functions, the relationship between nonlinear scattering function and other nonlinear network function and the relationship between nonlinear network function of the network link and the sub-network. The results of them have laid a theoretical basis for the proper use of this model.
Then, it is developed the extraction system of nonlinear scattering function based on the digital oscilloscope and the calibration methods of system errors. This system provided the experiment conditions to the extraction of nonlinear scattering function model of power devices. Nonlinear scattering function of power FETs and diodes can be tested and the results accurately reflected the nonlinear variation with the increase of input power to the power devices.
Links and differences of the model of nonlinear scattering function and the model of parameters X are proposed by Agilent recently are analyzed. Through the comparison of the description、characteristic、extraction methods and the applications between them, it provides a useful reference and comparison to the study and application of nonlinear scattering function.
Based on the test data of nonlinear scattering function of power devices, the new machine learning method of support vector machine is used to predict the model of nonlinear scattering function and the good predict accuracy is got. Then, it is summarized that the affection to the modeling accuracy from the selection of the support vector machine, kernel function and other parameters.
Last, the model of nonlinear scattering function of FETs and Schottky barrier diodes are used to design the microwave power amplifier and frequency multiplier. It is verified that the applicability of the large-signal model.
Compare with the traditional large-signal model, the advantages of this black-box model can be generalized. The first is the compatibility with the parameters S at small signal. Through this, the unified microwave circuit model can be established between the working conditions of large-signal and small-signal. Different from the foreign NNMS(Nonlinear network measurement system), the second is the time domain measurement method based on the digital oscilloscope. This method is simple and convenient. Adopting the common test fixture, a unified method can be used to exact the nonlinear scattering function model. The last, the model is frequency-domain black-box model, it can be extracted which is not needed to know the structure of microwave semiconductor devices and circuits in advance. The model not only contains some of the intrinsic part, and the affection of parasitic parameters is included in this model. It can be used in the design of microwave circuits conveniently and accurately.
It can be proved that this new large-signal model of microwave devices can improve the design efficiency, improve design accuracy and performance effectively.
In addition ,some problems, such as how to improve the consistency problem to the phase of nonlinear scattering function; modeling accuracy is not high to the part of nonlinear scattering function using support vector machine and how to design a unified large signal CAD(Computer aided design) model with the model of nonlinear scattering function, still requires in-depth study and explore.
引文
[1]章荣庆.微波电子线路.西安:西安电子科技大学出版社,1996.4-6
[2]Peter H Aaen,Jaime A.Pla,John Wood.鲍景富等译.射频微波功率场效应管的建模与特征.北京:电子工业出版社,2009,1-3
[3]刘祖深.微波毫米波测试仪器技术的新进展.电子测量与仪器学报.2009,23(3).1-8
[4]王卓鹏,高国成等GaAs MESFET管大信号建模研究.系统工程与电子技术.2002,24 (7 ). 123-125
[5]易卿武,于洪喜功率GaAs MESFET大信号建模.空间电子技术.2001,3 .15-22
[6]Schreurs D, Verspecht J, Nauwelaers B. Direct Extraction Of The Non-linear Model For Two-port Devices From Vectorial Non-linear Network Analyzer Measurements. 27th European Microwave Conference and Exhibition.1997,2. 921– 926
[7]林茂六,许洪光.超越S参数的大信号射频网络分析测试技术概述.测控技术.2003,22(3).1-4
[8]林茂六,郝绍杰,张亦驰.非线性矢量网络分析仪的原理和绝对校准技术.电子测量与仪器学报增刊.2009,23.136-143
[9]Q.J.Zhang,Kuldip C Gupta,Vigay K.Devabhaktuni.Artificial Neural Networks for RF and Microwave Design-from Theory to Pratice.IEEE Tran. on Microwave Theory to Pratice.2003,51(4).1339-1350
[10]Verspecht J, Large-signal network analysis. IEEE Microwave Magazine.2005,6(4).82-92
[11] D.E.Root etal.Broad-Band Poly-Harmonic Distortion(PHD)Behavioral Models From Fast Automated Simulations and Large-Signal Vectorial Network Measurements. IEEE Trans.MTT.2005,11.53(11).3656-3664.
[12]Vladimir N Vapnik,张学工译.统计学习理论的本质.北京:清华大学出版社,2009.9
[13]张学工.关于统计学习理论与支持向量机.自动化学报.2000,26(1).32-42.
[14]Cortes C,Vapnik V.Support-vector networks.Machine Learning.1995,20.273-297
[15]王惠功.非线性微波毫米波电路的分析和设计.北京:北京邮电学院出版社,1991.3-19
[16]王家礼.微波电路CAA与CAD.西安:西安电子科技大学出版社,2003
[17]倪峰.微波非线性散射函数仿真技术的研究.西安电子科技大学硕士论文,2008,1.
[18]高建军.场效应晶体管射频微波建模技术.北京:电子工业出版社,2007.236-245.
[19]吴万春,梁昌洪.微波网络及其应用.北京:国防工业出版社,1979.16-25
[20]Verspecht J, Dylan F,Williams and Dominique Schreurs. Linearization oflarge-signal scattering parameter. IEEE Transactions on Microwave Theory and Techniques. 2005,4. 53 .1369 - 1376
[21]Bossche Marc Vanden, Verbeyst Frans, Verspecht J. The Three Musketeers of Large Signal RF and Microwave Design-Measurement, Modeling and CAE . 53rd ARFTG Conference Digest-Spring 1999,35 .1– 8
[22]Verspecht J, Verbeyst Frans, vanden Bossche Marc.Network Analysis Beyond S-parameters: Characterizing and Modeling Component Behaviour under Modulated Large-Signal Operating Conditions. 56th ARFTG Conference Digest-Fall.2000,38. 1– 4
[23]J Verspecht ,P Van Esch. Accurately characterizing hard nonlinear behavior of microwave components with the nonlinear network measurement systems: Introducing Nonlinear scattering functions.Proceedings of 5th international workshop on Integrated Nonlinear Microwave and Millimeter Circuits.US:INMMC,1998
[24]J Verspecht , M Vanden Bossche F Verbeyst. Characterizing components under large signal excitation: Defining sensible‘Large signal S-parameters’?! 49th ARFTG CONFERENCE DIGEST, 1997 .1-10
[25] J Verspecht .Describing functions can better model hard nonlinearities in the frequency domain than the volterra theory .Dcctoral Dissertation Annex,Vrije University Brussel , Brussels,Belgium,1995
[26]孙璐,王家礼.一种微波非线性电路建模新方法.西安电子科技大学学报,2009, 36(2).294-297
[27] Sun Lu ,Wang Jiali etal.Definition and measurement technique of nonlinear network function to microwave circuit.2008 International Conference on Microwave and Millimeter Wave Technology Proceedings 2008,1:253-256
[28]林茂六,孙洪剑,姜靖.消除相位模糊的谐波相关复信号比值计算方法.电子学报,2006,6 34(6).1130-1133
[29]Schreurs Dominique,Verspecht J. Theoretical and experimental assessment of the non-linear scattering functions for the CAD of non-linear microwave circuits.2001 GAAS conference,1. 1-5
[30] J Verspecht.Black-box modelling of power transistors in the frequency domain.56th ARFTG Conference proceedings .2000,12 .1-7
[31]林茂六,于海雁.一种新的大信号射频功率器件建模方法.电子学报.2003,9,31(9). 1320-1322
[32]孙璐,王家礼.表征微波非线性网络参数的研究.微波毫米波测试仪器基础研究学术交流会, 2007,1.196-201
[33]孙璐,王家礼.非线性散射函数特性及其用于网络连接的研究.微波毫米波测试仪器基础研究学术交流会, 2007,1.202-208
[34]孙璐,王家礼.微波非线性电路测量和设计方法的新进展.微波毫米波测试仪器基础研究学术交流会, 2007,1.186-190
[35]郭生平,孙璐,王家礼.非线性散射函数的提取方法研究.非线性建模与分析研讨会2008,1.70-77
[36]Verspecht, J, Root D E,Wood J,Cognata A.Broad-band, multi-harmonic frequency domain behavioral models from automated large-signal vectorial network measurements Microwave Symposium Digest, 2005 IEEE MTT-S International.
[37] Verspecht, J. Broadband sampling oscilloscope characterization with the“nose-to-nose”calibration procedure: a theoretical analysis Instrumentation and Measurement Technology Conference, 1994,2 526– 529
[38]Barataud, D,Mallet A, Campovecchio M. Measurements of time domain voltage/current waveforms at R.F. and microwave frequencies for the characterization of nonlinear devices. IEEE Instrumentation and Measurement Technology Conference, 1998, 2. 1006– 1010
[39] Sun Lu ,Wang Jiali,Wang shan etal. Large-signal modeling Method for power FETs and Diodes .Journal of Semiconductors. 2009,30(6).064003-4.
[40]刘金现.通用测试夹具设计.微波毫米波测试仪器基础研究学术交流会, 2007,1:182-185
[41]Van den Broeck, Verspecht J. Calibrated vectorial nonlinear-network analyzers. IEEE MTT-S International Microwave Symposium Digest, 1994,2. 1069– 1072
[42]Schreurs D, Verspecht J,Nauwelaers. Direct Extraction of the Non-Linear Model for Two-Port Devices from Vectorial Non-Linear Network Analyzer Measurements. 27th European Microwave Conference, 1997.2 .921 - 926
[43]Scott J,Behnia B, Bossche M V.Removal of cable and connector dispersion in time-domain waveform measurements on 40 Gb integrated circuits. 2002 IEEE MTT-S International Microwave Symposium Digest, 2002,3 .1669 - 1672
[44] El Yaagoubi, M Neveux, G, Barataud D. Accurate phase measurements of broadband Multitone signals using a specific configuration of a Large Signal Network Analyzer. IEEE MTT-S International Microwave Symposium Digest, 2006. 1448– 1451
[45] D D Amore, P Maffezzoni, M Pillan. A Newton-Powell Modification Algorithm For Harmonic Balance-Based Circuit Analysis. IEEE Trans. on Circuits and Systems. 1994, 41(2). 177-180
[46]Chao-Ren Chang, Michael B Steer. Frequency-Domain Nonlinear Microwave Circuit Simulation Using the Arithmetic Operator Method. IEEE Trans. on MTT. 1990, 1139-1143
[47]曾学刚.微波非线性电路理论研究的新进展.电子学报.1996,6,6-10
[48]张祖舜,沈灿.用算术运算法分析多频率激励微波非线性电路.电子学报.1997,2.
[49]汪连栋,袁乃昌,王国玉.GaAs MESFET微波非线性电路全频域谐波平衡分析.雷达与对抗.1999,2
[50]王磊,杨红.射频电路设计技术.北京:电子工业出版社,2007,155-176
[51]Verspecht J,D Root.Polyharmonic Distortion Modeling.IEEE Microwave Magazine,2006,7(3).44-57
[52]孙璐,王家礼,陈晓龙.两种基于测量的微波非线性电路频域黑箱模型.电子测量与仪器学报2009,23(10).19-24
[53]Verspecht J, Horn, J, Betts L, Gunyan, D. Extension of X-parameters to include long-term dynamic memory effects. Microwave Symposium Digest, 2009. MTT '09. IEEE MTT-S International.2009, 741– 744
[54] Verspecht, J,Gunyan, D,Horn J. Multi-tone, Multi-port, and Dynamic Memory Enhancements to PHD Nonlinear Behavioral Models from Large-signal Measurements and Simulations. Microwave Symposium, 2007. IEEE/MTT-S International,2007, 969– 972
[55]华晓杰,林茂六,孙洪剑.多谐波失真模型应用研究.电讯技术.2007,47(3).19-23
[56]孙璐,王家礼.Agilent非线性X参数和非线性散射函数之比较.非线性建模与分析研讨会2008,1.61-69
[57]Lardeux C,Frison P.-L, Tison C, Souyris.Support Vector Machine for Multifrequency SAR Polarimetric Data Classification. IEEE Transactions on Geoscience and Remote Sensing,2009,47(12). 4143– 4152
[58] Ong, C.-J, Shao S, Yang J. An Improved Algorithm for the Solution of the Regularization Path of Support Vector Machine. IEEE Transactions on Neural Networks, 2010,21(3).451 - 462
[59]Lardeux, C, Frison, P.-L, Tison, C. Support Vector Machine for Multifrequency SAR Polarimetric Data Classification. IEEE Transactions on Geoscience and Remote Sensing.2009 7(12). 4143– 4152
[60]Zhi-biao Shi, Yang Li, Yun-feng Song. Fault Diagnosis of Transformer Based on Quantum-Behaved Particle Swarm Optimization-Based Least Squares Support Vector Machines. International Conference on Information Engineering and Computer Science, 2009,1 .1– 4
[61]Hang Xie, Yuhe Liao, Hao Tang. Reliable Prediction System Based on Support Vector Regression with Genetic Algorithms, Fifth International Conference on Natural Computation, 2009,1.552– 555
[62] Liu Xiaoli, Wang Jiali, Sun lu etal.Study of modeling for microwave high-powerdevices based on artificial neural networks.2008 International Conference on Microwave and Millimeter Wave Technology Proceedings 2008,1:261-264
[63]袁小芳,王耀南,孙炜.一种用于RBF神经网络的支持向量机与BP的混合算法.湖南大学学报.2005,6 ,32(3).88-92
[64]李芳芳,赵英凯,颜昕.基于Matlab的最小二乘支持向量机的工具箱及其应用.计算机应用.2006,12,26.358-360
[65]徐军辉,汪立新,钱培贤.基于最小二乘支持向量机的小样本建模方法研究.航天控制.2008,2 26(1).8-12
[66]董春曦,饶鲜,杨绍全.支持向量机参数选择方法研究.系统工程与电子技术.2004,8 .26(8).1117-1120
[67]郭丽娟,孙世宇,段修生.支持向量机及核函数研究.科学技术与工程.2008,1.8(2).487-490
[68]王华忠,俞金寿.核函数方法及其模型选择.江南大学学报.2006,8.5(4).500-504
[69]王凯,张永祥,姚晓山.支持向量机惩罚参数的自适应调整方法.计算机工程与应用.2008 ,44(26).45-47
[70]张爱华,于忠党.基于支持向量机回归的放大器性能评价研究.仪器仪表学报,2008,29(3).618-622
[71]华晓杰,林茂六.基于支持向量机的大信号射频功率器件特征建模.吉林大学学报.2008,38(1).185-189.
[72]孙璐,王家礼,陈晓龙.基于支持向量机的微波功率FET大信号建模方法的研究.电子测量与仪器学报增刊, 2009,23.189-193
[73]王子宇,张肇仪,徐承和等.射频电路设计理论与应用.北京:电子工业出版社,2008.
[74]Steve C Cripps. Advanced Techniques in RF Power Amplifier Design. Boston London: Artech House, 2002.
[75]P B Kenington. High-linearity RF Amplifier Design. Boston London: Artech House, 2000
[76]李润旗,李国定,陈兆清.微波电路CAD软件应用技术.北京.国防工业出版社.
[77]龚建. ADS软件对倍频器的嵌入式电特性仿真.国外电子元器件.2004,1.27-29
[78]徐振宇,钱澄.阶跃恢复二极管倍频器的设计.电子器件.2005.3 28(1).125-127
[79]张永鸿,唐小宏,樊勇.微波PIN二极管倍频器研究.固体电子学研究与进展.2007,11,27(4) .498-502
[80]朱志勇,王积勤.微波倍频器的发展与设计.制导与引信.2003,9.24(3).46-50
[81]黄绣江,唐宗熙,杨涛.Ka波段二倍频器的研究与设计.电子科技大学学报.2002,10.31(5)
[82]朱志勇.微波倍频器的设计.微电子与基础产品.2003,20(9).61-64
[2]Peter H Aaen,Jaime A.Pla,John Wood.鲍景富等译.射频微波功率场效应管的建模与特征.北京:电子工业出版社,2009,1-3
[3]刘祖深.微波毫米波测试仪器技术的新进展.电子测量与仪器学报.2009,23(3).1-8
[4]王卓鹏,高国成等GaAs MESFET管大信号建模研究.系统工程与电子技术.2002,24 (7 ). 123-125
[5]易卿武,于洪喜功率GaAs MESFET大信号建模.空间电子技术.2001,3 .15-22
[6]Schreurs D, Verspecht J, Nauwelaers B. Direct Extraction Of The Non-linear Model For Two-port Devices From Vectorial Non-linear Network Analyzer Measurements. 27th European Microwave Conference and Exhibition.1997,2. 921– 926
[7]林茂六,许洪光.超越S参数的大信号射频网络分析测试技术概述.测控技术.2003,22(3).1-4
[8]林茂六,郝绍杰,张亦驰.非线性矢量网络分析仪的原理和绝对校准技术.电子测量与仪器学报增刊.2009,23.136-143
[9]Q.J.Zhang,Kuldip C Gupta,Vigay K.Devabhaktuni.Artificial Neural Networks for RF and Microwave Design-from Theory to Pratice.IEEE Tran. on Microwave Theory to Pratice.2003,51(4).1339-1350
[10]Verspecht J, Large-signal network analysis. IEEE Microwave Magazine.2005,6(4).82-92
[11] D.E.Root etal.Broad-Band Poly-Harmonic Distortion(PHD)Behavioral Models From Fast Automated Simulations and Large-Signal Vectorial Network Measurements. IEEE Trans.MTT.2005,11.53(11).3656-3664.
[12]Vladimir N Vapnik,张学工译.统计学习理论的本质.北京:清华大学出版社,2009.9
[13]张学工.关于统计学习理论与支持向量机.自动化学报.2000,26(1).32-42.
[14]Cortes C,Vapnik V.Support-vector networks.Machine Learning.1995,20.273-297
[15]王惠功.非线性微波毫米波电路的分析和设计.北京:北京邮电学院出版社,1991.3-19
[16]王家礼.微波电路CAA与CAD.西安:西安电子科技大学出版社,2003
[17]倪峰.微波非线性散射函数仿真技术的研究.西安电子科技大学硕士论文,2008,1.
[18]高建军.场效应晶体管射频微波建模技术.北京:电子工业出版社,2007.236-245.
[19]吴万春,梁昌洪.微波网络及其应用.北京:国防工业出版社,1979.16-25
[20]Verspecht J, Dylan F,Williams and Dominique Schreurs. Linearization oflarge-signal scattering parameter. IEEE Transactions on Microwave Theory and Techniques. 2005,4. 53 .1369 - 1376
[21]Bossche Marc Vanden, Verbeyst Frans, Verspecht J. The Three Musketeers of Large Signal RF and Microwave Design-Measurement, Modeling and CAE . 53rd ARFTG Conference Digest-Spring 1999,35 .1– 8
[22]Verspecht J, Verbeyst Frans, vanden Bossche Marc.Network Analysis Beyond S-parameters: Characterizing and Modeling Component Behaviour under Modulated Large-Signal Operating Conditions. 56th ARFTG Conference Digest-Fall.2000,38. 1– 4
[23]J Verspecht ,P Van Esch. Accurately characterizing hard nonlinear behavior of microwave components with the nonlinear network measurement systems: Introducing Nonlinear scattering functions.Proceedings of 5th international workshop on Integrated Nonlinear Microwave and Millimeter Circuits.US:INMMC,1998
[24]J Verspecht , M Vanden Bossche F Verbeyst. Characterizing components under large signal excitation: Defining sensible‘Large signal S-parameters’?! 49th ARFTG CONFERENCE DIGEST, 1997 .1-10
[25] J Verspecht .Describing functions can better model hard nonlinearities in the frequency domain than the volterra theory .Dcctoral Dissertation Annex,Vrije University Brussel , Brussels,Belgium,1995
[26]孙璐,王家礼.一种微波非线性电路建模新方法.西安电子科技大学学报,2009, 36(2).294-297
[27] Sun Lu ,Wang Jiali etal.Definition and measurement technique of nonlinear network function to microwave circuit.2008 International Conference on Microwave and Millimeter Wave Technology Proceedings 2008,1:253-256
[28]林茂六,孙洪剑,姜靖.消除相位模糊的谐波相关复信号比值计算方法.电子学报,2006,6 34(6).1130-1133
[29]Schreurs Dominique,Verspecht J. Theoretical and experimental assessment of the non-linear scattering functions for the CAD of non-linear microwave circuits.2001 GAAS conference,1. 1-5
[30] J Verspecht.Black-box modelling of power transistors in the frequency domain.56th ARFTG Conference proceedings .2000,12 .1-7
[31]林茂六,于海雁.一种新的大信号射频功率器件建模方法.电子学报.2003,9,31(9). 1320-1322
[32]孙璐,王家礼.表征微波非线性网络参数的研究.微波毫米波测试仪器基础研究学术交流会, 2007,1.196-201
[33]孙璐,王家礼.非线性散射函数特性及其用于网络连接的研究.微波毫米波测试仪器基础研究学术交流会, 2007,1.202-208
[34]孙璐,王家礼.微波非线性电路测量和设计方法的新进展.微波毫米波测试仪器基础研究学术交流会, 2007,1.186-190
[35]郭生平,孙璐,王家礼.非线性散射函数的提取方法研究.非线性建模与分析研讨会2008,1.70-77
[36]Verspecht, J, Root D E,Wood J,Cognata A.Broad-band, multi-harmonic frequency domain behavioral models from automated large-signal vectorial network measurements Microwave Symposium Digest, 2005 IEEE MTT-S International.
[37] Verspecht, J. Broadband sampling oscilloscope characterization with the“nose-to-nose”calibration procedure: a theoretical analysis Instrumentation and Measurement Technology Conference, 1994,2 526– 529
[38]Barataud, D,Mallet A, Campovecchio M. Measurements of time domain voltage/current waveforms at R.F. and microwave frequencies for the characterization of nonlinear devices. IEEE Instrumentation and Measurement Technology Conference, 1998, 2. 1006– 1010
[39] Sun Lu ,Wang Jiali,Wang shan etal. Large-signal modeling Method for power FETs and Diodes .Journal of Semiconductors. 2009,30(6).064003-4.
[40]刘金现.通用测试夹具设计.微波毫米波测试仪器基础研究学术交流会, 2007,1:182-185
[41]Van den Broeck, Verspecht J. Calibrated vectorial nonlinear-network analyzers. IEEE MTT-S International Microwave Symposium Digest, 1994,2. 1069– 1072
[42]Schreurs D, Verspecht J,Nauwelaers. Direct Extraction of the Non-Linear Model for Two-Port Devices from Vectorial Non-Linear Network Analyzer Measurements. 27th European Microwave Conference, 1997.2 .921 - 926
[43]Scott J,Behnia B, Bossche M V.Removal of cable and connector dispersion in time-domain waveform measurements on 40 Gb integrated circuits. 2002 IEEE MTT-S International Microwave Symposium Digest, 2002,3 .1669 - 1672
[44] El Yaagoubi, M Neveux, G, Barataud D. Accurate phase measurements of broadband Multitone signals using a specific configuration of a Large Signal Network Analyzer. IEEE MTT-S International Microwave Symposium Digest, 2006. 1448– 1451
[45] D D Amore, P Maffezzoni, M Pillan. A Newton-Powell Modification Algorithm For Harmonic Balance-Based Circuit Analysis. IEEE Trans. on Circuits and Systems. 1994, 41(2). 177-180
[46]Chao-Ren Chang, Michael B Steer. Frequency-Domain Nonlinear Microwave Circuit Simulation Using the Arithmetic Operator Method. IEEE Trans. on MTT. 1990, 1139-1143
[47]曾学刚.微波非线性电路理论研究的新进展.电子学报.1996,6,6-10
[48]张祖舜,沈灿.用算术运算法分析多频率激励微波非线性电路.电子学报.1997,2.
[49]汪连栋,袁乃昌,王国玉.GaAs MESFET微波非线性电路全频域谐波平衡分析.雷达与对抗.1999,2
[50]王磊,杨红.射频电路设计技术.北京:电子工业出版社,2007,155-176
[51]Verspecht J,D Root.Polyharmonic Distortion Modeling.IEEE Microwave Magazine,2006,7(3).44-57
[52]孙璐,王家礼,陈晓龙.两种基于测量的微波非线性电路频域黑箱模型.电子测量与仪器学报2009,23(10).19-24
[53]Verspecht J, Horn, J, Betts L, Gunyan, D. Extension of X-parameters to include long-term dynamic memory effects. Microwave Symposium Digest, 2009. MTT '09. IEEE MTT-S International.2009, 741– 744
[54] Verspecht, J,Gunyan, D,Horn J. Multi-tone, Multi-port, and Dynamic Memory Enhancements to PHD Nonlinear Behavioral Models from Large-signal Measurements and Simulations. Microwave Symposium, 2007. IEEE/MTT-S International,2007, 969– 972
[55]华晓杰,林茂六,孙洪剑.多谐波失真模型应用研究.电讯技术.2007,47(3).19-23
[56]孙璐,王家礼.Agilent非线性X参数和非线性散射函数之比较.非线性建模与分析研讨会2008,1.61-69
[57]Lardeux C,Frison P.-L, Tison C, Souyris.Support Vector Machine for Multifrequency SAR Polarimetric Data Classification. IEEE Transactions on Geoscience and Remote Sensing,2009,47(12). 4143– 4152
[58] Ong, C.-J, Shao S, Yang J. An Improved Algorithm for the Solution of the Regularization Path of Support Vector Machine. IEEE Transactions on Neural Networks, 2010,21(3).451 - 462
[59]Lardeux, C, Frison, P.-L, Tison, C. Support Vector Machine for Multifrequency SAR Polarimetric Data Classification. IEEE Transactions on Geoscience and Remote Sensing.2009 7(12). 4143– 4152
[60]Zhi-biao Shi, Yang Li, Yun-feng Song. Fault Diagnosis of Transformer Based on Quantum-Behaved Particle Swarm Optimization-Based Least Squares Support Vector Machines. International Conference on Information Engineering and Computer Science, 2009,1 .1– 4
[61]Hang Xie, Yuhe Liao, Hao Tang. Reliable Prediction System Based on Support Vector Regression with Genetic Algorithms, Fifth International Conference on Natural Computation, 2009,1.552– 555
[62] Liu Xiaoli, Wang Jiali, Sun lu etal.Study of modeling for microwave high-powerdevices based on artificial neural networks.2008 International Conference on Microwave and Millimeter Wave Technology Proceedings 2008,1:261-264
[63]袁小芳,王耀南,孙炜.一种用于RBF神经网络的支持向量机与BP的混合算法.湖南大学学报.2005,6 ,32(3).88-92
[64]李芳芳,赵英凯,颜昕.基于Matlab的最小二乘支持向量机的工具箱及其应用.计算机应用.2006,12,26.358-360
[65]徐军辉,汪立新,钱培贤.基于最小二乘支持向量机的小样本建模方法研究.航天控制.2008,2 26(1).8-12
[66]董春曦,饶鲜,杨绍全.支持向量机参数选择方法研究.系统工程与电子技术.2004,8 .26(8).1117-1120
[67]郭丽娟,孙世宇,段修生.支持向量机及核函数研究.科学技术与工程.2008,1.8(2).487-490
[68]王华忠,俞金寿.核函数方法及其模型选择.江南大学学报.2006,8.5(4).500-504
[69]王凯,张永祥,姚晓山.支持向量机惩罚参数的自适应调整方法.计算机工程与应用.2008 ,44(26).45-47
[70]张爱华,于忠党.基于支持向量机回归的放大器性能评价研究.仪器仪表学报,2008,29(3).618-622
[71]华晓杰,林茂六.基于支持向量机的大信号射频功率器件特征建模.吉林大学学报.2008,38(1).185-189.
[72]孙璐,王家礼,陈晓龙.基于支持向量机的微波功率FET大信号建模方法的研究.电子测量与仪器学报增刊, 2009,23.189-193
[73]王子宇,张肇仪,徐承和等.射频电路设计理论与应用.北京:电子工业出版社,2008.
[74]Steve C Cripps. Advanced Techniques in RF Power Amplifier Design. Boston London: Artech House, 2002.
[75]P B Kenington. High-linearity RF Amplifier Design. Boston London: Artech House, 2000
[76]李润旗,李国定,陈兆清.微波电路CAD软件应用技术.北京.国防工业出版社.
[77]龚建. ADS软件对倍频器的嵌入式电特性仿真.国外电子元器件.2004,1.27-29
[78]徐振宇,钱澄.阶跃恢复二极管倍频器的设计.电子器件.2005.3 28(1).125-127
[79]张永鸿,唐小宏,樊勇.微波PIN二极管倍频器研究.固体电子学研究与进展.2007,11,27(4) .498-502
[80]朱志勇,王积勤.微波倍频器的发展与设计.制导与引信.2003,9.24(3).46-50
[81]黄绣江,唐宗熙,杨涛.Ka波段二倍频器的研究与设计.电子科技大学学报.2002,10.31(5)
[82]朱志勇.微波倍频器的设计.微电子与基础产品.2003,20(9).61-64