跳频码序列建模与预测研究

【作者】 涂靖；

【导师】 魏平；

【作者基本信息】 电子科技大学 ， 信号与信息处理， 2008， 硕士

【摘要】 随着通信领域的频率竞争愈演愈烈,惯用的定频通信受到了严重威胁。为了保证己方正常可靠的通信,一种抗干扰通信体制——跳频通信系统应运而生。跳频通信技术具有优良的抗干扰性能和多址组网性能,不但在军事通信中得到了广泛的应用,如美国的SINCGARS系列超短波跳频电台和联合战术信息分发系统JTIDS,而且与我们的日常生活密切相关,在民用移动通信中也得到了广泛应用,如GSM、HomeRF(家庭射频)、Bluetooth(蓝牙)中都应用了跳频技术。跳频通信通过躲避干扰的方法来达到优良的抗干扰性能,仅当干扰信号频率与跳频信号频率相同时,才能形成干扰,因此对其干扰和抗干扰的研究成为通信电子战的一个重要部分。而用来控制载波频率跳变的序列对跳频通信系统的抗干扰能力有着决定性的影响,所以对跳频序列的研究成为当前国内外的一个热点。本文主要研究跳频码序列的建模与预测,即对跳频序列的历史观察值x1 , x2 , , xn ?1 ,xn进行建模,预测未来时刻xn + k( k> 0)的值。具体工作内容如下:1)阐述了常规跳频通信系统的原理,对系统各个组成部分的作用作了详细介绍,综述了跳频组网和跳频同步方法,并对部分常见跳频序列的设计进行了研究,最后对常规跳频通信系统进行了仿真实验。2)对基于移位寄存器产生的跳频码进行了分析,总结出此类跳频码的构造规律以达到预测的目的。本文提出了一种专门针对基于移位寄存器构造的跳频序列的预测方法,在已知很少频点时即可进行预测,预测速度快,有效预测率高。3)运用非线性预测方法对常见跳频码进行预测。文中主要运用BP神经网络、RBF神经网络、Bernstein多项式、Volterra自适应滤波器四种预测模型进行预测,通过理论分析和仿真实验对各种预测模型的性能进行了比较分析,并通过半实物仿真实验验证了各种预测算法的有效性。4)对混沌时间序列进行了多步和多模式预测。本文研究了两种多步预测方法,首先从理论上分析了两种方法的优劣,然后再用上述四种非线性预测模型对两种方法进行了仿真实验,并在以上非线性预测模型的基础上进行了多模式预测,仿真结果表明多模式预测能提高序列的有效预测率。更多还原

【Abstract】 As the competition in frequency become more and more intensively in communication domain, the communication with fixed frequency suffers severe threats. In order to guarantee our communication, a new anti-Jamming communication system emerges, namely Frequency Hopping Communication System(FHCS). Duing to its outstanding counter-interference capability and Multiple Access properties, FHCS not only becomes popular in military communication, such as SINCGARS and JTIDS, but also is widely used in the civil mobile communications, including GSM, HomeRF and Bluetooth.FH communication has anti-scout and anti-jamming capability by escaping the jamming. It can work when the frequency hops irregularly and quickly, which can’t be intercepted, recognised and interfered easily. It is interfered only when the signal frequency equals FH signal frequency. So the researches on anti-jamming and jamming of FHCS are a very important part of electronic warfare.The FH sequences which are used to control the frequency affects the anti-jamming capability crucially.In order to improve the anti-jamming and jamming capability of FHCS, the main work is to predict FH Sequence in the dissertation, including the following several aspects.1) The basic principle of FHCS’s and FH sequences is presented, the components of the system are described detailedly, FH mutiple access and synchronization methods are summarized, the design for FH sequences is studied. At last, computer simulations are done.2) The construction principle of some familiar FH Sequences is analyzed. A novel predicting method aiming at the FH Sequences producted by shift-register is proposed. Computer simulations show the feasibility and efficiency of the method.3) Several nonlinear predicting methods are analyzed and studied. BP Neural Network, RBF Neural Network, Bernstein polynomial, Volterra adaptive filter are used for predicting. The advantages and disadvantages of each predicting method in predicting different kinds of sequences are analyzed. 4) Multi-step and multi-mode forecasting for chaotic time series are researched. Compare the performance of two multi-step forecasting methods according to theory analysis and experimental results. Multi-mode forecasting based on the above four nonlinear predicting methods is made, and the simulation results indicate that effective forecasting probability can be increased.更多还原