基于神经网络的无线信道的辨识与预测
|
摘要
本论文首先分析了无线信道的传输特性，并深入研究了移动信道模型。由于传统信道模型被认为是一个线性的非时变的信道，虽然，后来一些科研工作者又进行了深入的研究，把信道模型扩展到线性时变模型。但是，随着移动用户的增多、业务的飞速发展，频率复用、时间复用达到了它们模型的极限，此外，用户还要求有在线视频的服务。从而必然加剧信号传输的线性、非线性失真。这些粗糙的模型远远不能满足应用的需要，明显是过时的模型。用这些传统的信道模型去处理这些多信号、高速数据及多子载波传输信号，已经根本不能满足移动用户的需要，这就要求新的模型和新的信号处理方法去解决这一问题。我们提出基于神经网络的信号处理方法，由于神经网络的一些优点，使得该方法具有很好的应用前景。移动信道与电波传播，已有许多理论分析和现场实测的课题，并已得出有用的结果。其中有些给出精确的数学描述，另一些则给出统计模型。然而，由于移动信道的复杂性，目前仍有许多待研究的课题，我们不可能用单一的数学模型来描述所有的移动环境。
由于神经网络具有学习能力、容错能力或鲁棒性，可以实现输入到输出的非线性映射。因此，神经网络在非线性信号处理中越来越得到广泛的应用，其中一些算法还可以优化系统、预测信号，从而，可以检测、避开信号在信道中传输时发生的深衰落(deep fading)。预先知道接受信号、信道的一些先验知识，还可以降低数据传输误码率。优化网络的能力还可以提高数据传输速率。为了频率、时间的更大的共享，我们提出了用神经网络来辨识、预测这种非线性时变信道。论文提出的神经网络是目前很新颖的结构RNN网络结构，同时还提出用BP算法实现信号处理，把RNN(FRNN)的实时处理能力应用于通信系统之中，更是一个新颖的课题，论文还把神经网络融合到其它的一些模型中，提出组合预测的方法。比如，基于神经网络的ARMA模型的建立等。
总之，神经网络在无线通信中的应用，越来越受到重视和深入研究，势必使神经网络更深入融合到通信信号处理领域中，开创一个新的信号处理分支。
At first, This paper analyzes transmission characteristic of wireless channel, and deeply research mobile channel model. Since conventional channel model be considered as a linear time invariant channel, although, some research works continues to deeper research, and model linear time-varying channel. But, with the development of increasing mobile consumer and serves, frequency and time multiplexing maximize control mode in these models, including, consumes' online video serve. So these aggravate linear and nonlinear distortion of signal transmission. Now, these rough models can't meet application need, and they are obsolete by a 11 appearance. We apply these conventional to process these multi-signal, high-speed data and multi-carrier transmission signals, it can't meet mobile consumers' need at all. These requests need new model and signal processing to solve it. So we present based on neural network (ran) signal processing methods. Because of nn's many merits, this subject has a good application foreground. Mobile channel and electric wave transmission have had many subjects of principle analysis and field measurement, and attained useful results. In these subjects, some give accurate mathematic description, others give statistic models, many subjects still need to research now, for example, it's impossible to apply single mathematic model to describe all mobile environment.
Because nn have learning ability, fault-tolerant ability or robustness, and can implement nonlinear mapping input to output, nn obtain extensive use in the nonlinear signal processing, hereinto, nn can detect and avoid signal's deep fading in the wireless channel transmission, it can degrade data transmission error rate, and increase data transmission rate. For upward reasons, we present that nonlinear time-varying channel be identified by RNN. RNN is a new network architecture. At the same time, we also present that communication signal be processed by BP algorithm. Of course, it is a new subject. This paper also present that nn combine others prediction models, that is, combination prediction.
In a word, this paper presents that nn is applied to wireless communication
system, particularly, in the wireless channel. Signal processing based on nn is a new promising subject and different signal processing field.
由于神经网络具有学习能力、容错能力或鲁棒性，可以实现输入到输出的非线性映射。因此，神经网络在非线性信号处理中越来越得到广泛的应用，其中一些算法还可以优化系统、预测信号，从而，可以检测、避开信号在信道中传输时发生的深衰落(deep fading)。预先知道接受信号、信道的一些先验知识，还可以降低数据传输误码率。优化网络的能力还可以提高数据传输速率。为了频率、时间的更大的共享，我们提出了用神经网络来辨识、预测这种非线性时变信道。论文提出的神经网络是目前很新颖的结构RNN网络结构，同时还提出用BP算法实现信号处理，把RNN(FRNN)的实时处理能力应用于通信系统之中，更是一个新颖的课题，论文还把神经网络融合到其它的一些模型中，提出组合预测的方法。比如，基于神经网络的ARMA模型的建立等。
总之，神经网络在无线通信中的应用，越来越受到重视和深入研究，势必使神经网络更深入融合到通信信号处理领域中，开创一个新的信号处理分支。
At first, This paper analyzes transmission characteristic of wireless channel, and deeply research mobile channel model. Since conventional channel model be considered as a linear time invariant channel, although, some research works continues to deeper research, and model linear time-varying channel. But, with the development of increasing mobile consumer and serves, frequency and time multiplexing maximize control mode in these models, including, consumes' online video serve. So these aggravate linear and nonlinear distortion of signal transmission. Now, these rough models can't meet application need, and they are obsolete by a 11 appearance. We apply these conventional to process these multi-signal, high-speed data and multi-carrier transmission signals, it can't meet mobile consumers' need at all. These requests need new model and signal processing to solve it. So we present based on neural network (ran) signal processing methods. Because of nn's many merits, this subject has a good application foreground. Mobile channel and electric wave transmission have had many subjects of principle analysis and field measurement, and attained useful results. In these subjects, some give accurate mathematic description, others give statistic models, many subjects still need to research now, for example, it's impossible to apply single mathematic model to describe all mobile environment.
Because nn have learning ability, fault-tolerant ability or robustness, and can implement nonlinear mapping input to output, nn obtain extensive use in the nonlinear signal processing, hereinto, nn can detect and avoid signal's deep fading in the wireless channel transmission, it can degrade data transmission error rate, and increase data transmission rate. For upward reasons, we present that nonlinear time-varying channel be identified by RNN. RNN is a new network architecture. At the same time, we also present that communication signal be processed by BP algorithm. Of course, it is a new subject. This paper also present that nn combine others prediction models, that is, combination prediction.
In a word, this paper presents that nn is applied to wireless communication
system, particularly, in the wireless channel. Signal processing based on nn is a new promising subject and different signal processing field.
引文
[1] 张贤达，保铮，《通信信号处理》，国防工业出版社，2000
[2] Theodore S. Rappaport, "Wireless Communication Principles and Practice", Prentice Hall, 1996
[3] Simon Haykin, "Neural Networks=神经网络的综合基础：a comprehensive foundation", BeiJing: Tsinghua University Press, 2001
[4] John G. Proakis, "Digital Communication", New York, McGreawHill, 1995
[5] 朱洪波，《微蜂窝多径衰落信道中的无线电信号传播模式》，南京邮电学院学报，Vol．17，No．4，December 1997
[6] 赵荣黎，《数字移动通信传播特性预测及信道模型的研究》，北方交通大学学报，Vol．21，No．5，October 1997
[7] 赵荣黎，《数字蜂房移动通信系统》，电子工业出版社，1997
[8] Bello P A. "Characterization off randomly time-variant linear channels", IEEE Trans. Commun. Syst., 11: 360-393
[9] Proakis J G., "Digital Communications(3rd Edition)", McGraw Hill,1995
[10] Seshadri M., "Joint data and channel estimation using fast blind trellis search techniques", IEEE Trans. Commun., 1994, 42:1000-1011
[11] Crespo P M, Jimenez J., "Computer simulation of radio channels using a harmonic decomposition technique" IEEE Trans. Vehicular Technology,1995, 44: 414-419
[12] Bodson D, McClure G F, McConoughey S R., "Land-mobile communications engineering", New York: IEEE Press, 1984
[13] Jakes W C., "Microwave Communications", John Wiley, 1974
[14] Hoeher P., "A statistical discrete-time model for the WSSUS multipath channel", IEEE Trans. Vechcular Technology, 1992, 41: 461-467
[15] 郭梯云，杨家玮，李建东，《数字移动通信》，人民邮电出版社，1995
[16] 张成乾，张国强，《系统辨识与参数估计》，机械工业出版社，1985
[17] Papoulis A. Probability, "Random Variables and Stochastic Processes",
New York: McGraw-Hill, 1965
[18] Simon Haykin, "Neural network: a comprehensive foundation", New York: Maxwell Macmillan International, 1994
[19] Martin T. Hagan, Howard B. Demuth, Mark Beale, "Neural network design",PWS Pub., 1996
[20] Robert L. Harvey, "Neural network principles", Prentice Hall, 1994
[21] A. Cichocki, R. Unbehauen., "Neural networks for optimization and signal processing", Chichester: John Wiley&Sons, 1993
[22] 楼顺天，施阳，基于MATLAB的系统分析与设计[专著]-3，神经网络，1998．9
[23] Arun V. Holden, Vitaly I. Kryukov, "Neural network: theory and architecture", New York: Manchester University Press, 1998
[24] 王振龙，顾岚，《时间序列分析》，中国统计出版社，1999
[25] J. G. Taylor, "Neural networks and their applications", New York: UNICOM;Wiley, 1996
[26] 张大鹏，“神经网络系统设计方法=Neural Networks System Design Methodology”，北京：清华大学出版社，1996．5
[27] J. Stephen Judd, "Neural network design and the complexity of learning", Cambridge, Mass.: MIT Press, 1990
[28] S.M Pandit and Shien Ming Wu(1983), "Time Series and Analysis with Applications", Maclison John Wiley and Sons.
[29] Jamila Bakkoury, Daniel Roviras, Mounir Ghogho, Francis Castanie, "Adaptive MLSE receiver over rapidly fading channels", Signal Processing 80 (2000) pp 1347-1360
[30] 杜金观，项静怡，《时间序列分析-建模与预报》，安徽教育出版社，1990
[31] 翁文波，《预测论基础》，石油工业出版社，1984
[32] 张贤达，《现代信号处理》，清华大学出版社，1995
[33] George E. P. Box, Gwilym M. Jenkins, "Time series analysis: forecasting and control", Prentice-Hall, Inc, 1997
[34] M. K. Tsatsanis, Georgios B. Giannakis, and G. Zhou, "Estimation and
equalization of fading channels with random coefficients", Signal Processing, vol. 53, pp. 211-229, 1996
[35]Po-Rong Chang, Bor-Chin Wang, "Adaptive Decision Feedback Equalization for Digital Satellite Channels Using Multilayer Neural Networks", IEEE Journal on Selected Areas in Communications, Vol. 13,No. 2, pp. 316-324, Feb. 1995
[36]Spyros M eta., "Forecasting Methods and Applications", John Wiley & Sons, inc., 1983
[37]Rucy S T, George C T., "Consistent Estimates of Autoregressive Parameters and Extended Sample Autocorrelation Function for Stationary and Non-stationary ARMA Models", Journal of the American statistical Association,, March 1984, 79.
[38]Jerry B,, John C.,"Discrete-event System Simulation", Prentice-Hall,Inc. 1984.
[39]Akaike H., "A new look at the statistical model identification",IEEE Transactions on Automatic Control, 1974
[40]钟仪信等，《智能理论与技术-人工智能与神经网络》，人民邮电出版社，1992
[41]Steve B, Cyril O., "A Comparison of Box-Jenkins and Objective Methods for Determining the Order of Non-seasonal ARMA model", Journal of Forecasting, 1994, 13: 419-434
[42]Sudhakar M P, Shien-Ning Wu, "Time Series and System analysis With Applications", John Wiley & Sons, Inc., 1983
[43]焦李成，《非线性传递函数理论与应用》，西安电子科技大学出版社，1992
[44]曹建福，韩崇昭，方洋旺，《非线性系统理论及应用》，西安交通大学出版社，2001
[45]杨位钦，顾岚，《时间序列分析与动态数据建模》，北京理工大学出版社，1999
[46]S. Chen, G. J. Gibson, C. F. N. Cowan and P. M. Grant, "Adaptive Equalization of Finite Non-linear Channels Using Multilayer
Perceptrons", Signal Processing, Vol. 20, No. 2, pp. 107-119, June 1990
[47]KOH, T., Powers, E. J., "Second-order Volterra filtering and its application to nonlinear system identification", IEEE Trans., Dec. 1985, ASSP-33, (6), pp. 1445-1455
[48]K. Abend and B. D. Fritchman,"Statistical detection for communication channels with intersymbol interference", Proc. IEEE, Vol. 58,pp. 779-785, May 1970
[49]M.T. Ozden, A.H. Kayran and E. Panayirci, "Adaptive Volterra channal equalization with lattice orthogonalisation", IEE Proc.-Commun., Vol.145, No. 2, pp. 109-115, April 1998
[50]Bor-Sen Chen, Yih-Jinn Huang, Sin-Chung Chen, "Estimation and equalization of multipath Rician fading channels with stochastic tap coefficients", Signal Processing 68(1998) pp 43-57
[51]R. Steele. "Mobile Communications." Pentech Press, London, 1992.
[52]E. Baccarelli, R. Cusani, "Parameter identification of quasi-stationary Rayleigh-faded time-varying digital channels",Signal Processing 79(1999) pp1-13
[53]A. Jagoda and M. de Villepin, "Mobile Communications", New York: J. Wiley, 1993
[54]W.C. Jakes, Ed, "Microwave mobile Communication", NY: IEEE Press, 1993
[55]LIN T, et al., "Learning long-term dependencies in NARX recurrent neural networks", IEEE Transactions on Signal Processing, 1998, 46(8): 2207-2216
[56]Williams R J, Zipser D., "Gradient-based learning algorithms for recurrent networks and their computational complexity [A]", NJ: Lawrence ERLBAUM, 19945, Chapter 13
[2] Theodore S. Rappaport, "Wireless Communication Principles and Practice", Prentice Hall, 1996
[3] Simon Haykin, "Neural Networks=神经网络的综合基础：a comprehensive foundation", BeiJing: Tsinghua University Press, 2001
[4] John G. Proakis, "Digital Communication", New York, McGreawHill, 1995
[5] 朱洪波，《微蜂窝多径衰落信道中的无线电信号传播模式》，南京邮电学院学报，Vol．17，No．4，December 1997
[6] 赵荣黎，《数字移动通信传播特性预测及信道模型的研究》，北方交通大学学报，Vol．21，No．5，October 1997
[7] 赵荣黎，《数字蜂房移动通信系统》，电子工业出版社，1997
[8] Bello P A. "Characterization off randomly time-variant linear channels", IEEE Trans. Commun. Syst., 11: 360-393
[9] Proakis J G., "Digital Communications(3rd Edition)", McGraw Hill,1995
[10] Seshadri M., "Joint data and channel estimation using fast blind trellis search techniques", IEEE Trans. Commun., 1994, 42:1000-1011
[11] Crespo P M, Jimenez J., "Computer simulation of radio channels using a harmonic decomposition technique" IEEE Trans. Vehicular Technology,1995, 44: 414-419
[12] Bodson D, McClure G F, McConoughey S R., "Land-mobile communications engineering", New York: IEEE Press, 1984
[13] Jakes W C., "Microwave Communications", John Wiley, 1974
[14] Hoeher P., "A statistical discrete-time model for the WSSUS multipath channel", IEEE Trans. Vechcular Technology, 1992, 41: 461-467
[15] 郭梯云，杨家玮，李建东，《数字移动通信》，人民邮电出版社，1995
[16] 张成乾，张国强，《系统辨识与参数估计》，机械工业出版社，1985
[17] Papoulis A. Probability, "Random Variables and Stochastic Processes",
New York: McGraw-Hill, 1965
[18] Simon Haykin, "Neural network: a comprehensive foundation", New York: Maxwell Macmillan International, 1994
[19] Martin T. Hagan, Howard B. Demuth, Mark Beale, "Neural network design",PWS Pub., 1996
[20] Robert L. Harvey, "Neural network principles", Prentice Hall, 1994
[21] A. Cichocki, R. Unbehauen., "Neural networks for optimization and signal processing", Chichester: John Wiley&Sons, 1993
[22] 楼顺天，施阳，基于MATLAB的系统分析与设计[专著]-3，神经网络，1998．9
[23] Arun V. Holden, Vitaly I. Kryukov, "Neural network: theory and architecture", New York: Manchester University Press, 1998
[24] 王振龙，顾岚，《时间序列分析》，中国统计出版社，1999
[25] J. G. Taylor, "Neural networks and their applications", New York: UNICOM;Wiley, 1996
[26] 张大鹏，“神经网络系统设计方法=Neural Networks System Design Methodology”，北京：清华大学出版社，1996．5
[27] J. Stephen Judd, "Neural network design and the complexity of learning", Cambridge, Mass.: MIT Press, 1990
[28] S.M Pandit and Shien Ming Wu(1983), "Time Series and Analysis with Applications", Maclison John Wiley and Sons.
[29] Jamila Bakkoury, Daniel Roviras, Mounir Ghogho, Francis Castanie, "Adaptive MLSE receiver over rapidly fading channels", Signal Processing 80 (2000) pp 1347-1360
[30] 杜金观，项静怡，《时间序列分析-建模与预报》，安徽教育出版社，1990
[31] 翁文波，《预测论基础》，石油工业出版社，1984
[32] 张贤达，《现代信号处理》，清华大学出版社，1995
[33] George E. P. Box, Gwilym M. Jenkins, "Time series analysis: forecasting and control", Prentice-Hall, Inc, 1997
[34] M. K. Tsatsanis, Georgios B. Giannakis, and G. Zhou, "Estimation and
equalization of fading channels with random coefficients", Signal Processing, vol. 53, pp. 211-229, 1996
[35]Po-Rong Chang, Bor-Chin Wang, "Adaptive Decision Feedback Equalization for Digital Satellite Channels Using Multilayer Neural Networks", IEEE Journal on Selected Areas in Communications, Vol. 13,No. 2, pp. 316-324, Feb. 1995
[36]Spyros M eta., "Forecasting Methods and Applications", John Wiley & Sons, inc., 1983
[37]Rucy S T, George C T., "Consistent Estimates of Autoregressive Parameters and Extended Sample Autocorrelation Function for Stationary and Non-stationary ARMA Models", Journal of the American statistical Association,, March 1984, 79.
[38]Jerry B,, John C.,"Discrete-event System Simulation", Prentice-Hall,Inc. 1984.
[39]Akaike H., "A new look at the statistical model identification",IEEE Transactions on Automatic Control, 1974
[40]钟仪信等，《智能理论与技术-人工智能与神经网络》，人民邮电出版社，1992
[41]Steve B, Cyril O., "A Comparison of Box-Jenkins and Objective Methods for Determining the Order of Non-seasonal ARMA model", Journal of Forecasting, 1994, 13: 419-434
[42]Sudhakar M P, Shien-Ning Wu, "Time Series and System analysis With Applications", John Wiley & Sons, Inc., 1983
[43]焦李成，《非线性传递函数理论与应用》，西安电子科技大学出版社，1992
[44]曹建福，韩崇昭，方洋旺，《非线性系统理论及应用》，西安交通大学出版社，2001
[45]杨位钦，顾岚，《时间序列分析与动态数据建模》，北京理工大学出版社，1999
[46]S. Chen, G. J. Gibson, C. F. N. Cowan and P. M. Grant, "Adaptive Equalization of Finite Non-linear Channels Using Multilayer
Perceptrons", Signal Processing, Vol. 20, No. 2, pp. 107-119, June 1990
[47]KOH, T., Powers, E. J., "Second-order Volterra filtering and its application to nonlinear system identification", IEEE Trans., Dec. 1985, ASSP-33, (6), pp. 1445-1455
[48]K. Abend and B. D. Fritchman,"Statistical detection for communication channels with intersymbol interference", Proc. IEEE, Vol. 58,pp. 779-785, May 1970
[49]M.T. Ozden, A.H. Kayran and E. Panayirci, "Adaptive Volterra channal equalization with lattice orthogonalisation", IEE Proc.-Commun., Vol.145, No. 2, pp. 109-115, April 1998
[50]Bor-Sen Chen, Yih-Jinn Huang, Sin-Chung Chen, "Estimation and equalization of multipath Rician fading channels with stochastic tap coefficients", Signal Processing 68(1998) pp 43-57
[51]R. Steele. "Mobile Communications." Pentech Press, London, 1992.
[52]E. Baccarelli, R. Cusani, "Parameter identification of quasi-stationary Rayleigh-faded time-varying digital channels",Signal Processing 79(1999) pp1-13
[53]A. Jagoda and M. de Villepin, "Mobile Communications", New York: J. Wiley, 1993
[54]W.C. Jakes, Ed, "Microwave mobile Communication", NY: IEEE Press, 1993
[55]LIN T, et al., "Learning long-term dependencies in NARX recurrent neural networks", IEEE Transactions on Signal Processing, 1998, 46(8): 2207-2216
[56]Williams R J, Zipser D., "Gradient-based learning algorithms for recurrent networks and their computational complexity [A]", NJ: Lawrence ERLBAUM, 19945, Chapter 13
© 2004-2018 中国地质图书馆版权所有 京ICP备05064691号 京公网安备11010802017129号 地址：北京市海淀区学院路29号 邮编：100083 电话：办公室：(+86 10)66554848；文献借阅、咨询服务、科技查新：66554700 |