混沌动力系统参数辨识及其在保密通信中的应用研究

发布时间：2018-02-28 17:06

本文关键词： 混沌理论自适应参数辨识保密通信复杂网络　出处：《北京邮电大学》2014年博士论文　论文类型：学位论文

【摘要】：随着信息技术的飞速发展和大数据时代的到来,人们在享受网络带来更多便利的同时,越来越多的数据和信息在传输过程中也面临着监听、窃取、恶意篡改等威胁。信息安全问题成为这个时代影响国家安全和社会发展的重大问题,现有的保密通信方法面临着越来越多的挑战,研究新的保密通信方法来保护信息和数据的安全通信成为通信、计算机、信息等相关领域中的前沿和热点方向之一。混沌信号由于具有对初始条件的敏感性、非周期性、类噪声性、连续宽带频谱等固有属性,使其具有高度的不可预测性和天然的隐蔽性,特别适合作为保密通信中的传输载体。二十世纪九十年代初,混沌保密通信的想法一经问世,便成为信息安全领域研究的热点问题,国内外科学家和研究人员从信息的加密算法、抗攻击性、混沌器件等不同角度对混沌保密通信模型及其相关技术展开了深入研究。如今在大数据时代的影响下,混沌保密通信的研究方向也开始向着数字化、多用户、高性能、宽带无线等方面发展。通常情况下,混沌保密通信模型中的混沌系统存在着部分参数未知或全部参数未知的情况,这种未知性尽管会使得系统更难操作和控制,却会给利用这些参数进行信息调制的方法带来更高的安全性,提高了通信过程的保密性。为了更好地操作这类系统并实现信息的加密,可以将非线性控制理论中的自适应参数辨识技术引入带有未知参数的混沌系统,通过对系统状态或输出的不断测量,与预期目标进行比较,根据这个差值调整系统的结构或参数以实现控制的作用,周而复始直至达到最终目标。本论文研究了混沌系统的自适应参数辨识技术及其在保密通信中的应用,主要创新工作归纳如下： (1)本论文提出了一个自适应参数辨识中的新型判定条件——格拉姆矩阵条件,这个条件的提出建立了持续激励条件和线性无关条件之间的桥梁。持续激励条件或线性无关条件是自适应参数辨识中应用广泛的两个基本条件,但是它们之间的关系一直悬而未决。不同于这两个条件的计算方法,所提出的格拉姆矩阵条件通过计算矩阵的秩来判断未知参数的收敛情况,在某些情况下,能够降低计算持续激励条件或线性无关条件时的运算量,更具有实用性和高效性。 (2)本论文指出了近年来多篇自适应参数辨识文章中在设计辨识方法时的错误。传统线性无关条件的定义中缺少了对时间的限制,对于一个随时间变化的函数,无法通过普通的线性无关定义判断其相关性。本论文提出了持续激励条件、线性无关条件或格拉姆矩阵条件在长时间和有限时间情况下的相关定义,由此分析了混沌系统同步过程和判定条件持续时间长短对辨识结果的影响。除此之外,本论文提出了在噪声存在情况下参数辨识的方法,以及降低噪声干扰的措施。这部分中提出的判定条件和方法都可以应用在混沌保密通信方案中。 (3)本论文提出了一种基于自适应参数辨识的混沌通信模型,在一维标量混沌信号中加载了多个信息,实现了多信息的混沌保密通信传输。在这种通信模型中,原始信息首先被直接调制在发送端带有多个时滞的标量混沌系统的参数中,在接收端通过自适应参数辨识的方法对加载在混沌信号中的信息进行解调。这个由一维混沌系统构造的混沌通信模型携带的信息量比传统混沌通信模型携带的信息量多得多并且系统更加简单。接着,基于沃尔什函数构成的正交基,本论文将多路原始信息与沃尔什序列进行正交调制,再加载在发送端的混沌系统参数中,更大程度地提高了系统携带的信息总量。本论文所提出的混沌保密通信方案在混沌激光通信领域有潜在的应用前景。 (4)本论文提出了一种利用混沌同步暂态过程实现复杂动态网络系统结构辨识的方法。前面研究了两个混沌系统构成的混沌保密通信方案,实现了点对点的混沌通信传输。多个混沌通信系统相互通信就构成了复杂动态网络的结构。近年来研究复杂动态网络的文章普遍认为网络内部的同步现象阻碍拓扑结构辨识的完成。而本论文深入探索了网络同步的暂态过程,利用所提出的自适应参数辨识方法研究复杂网络内部节点间的拓扑连接,发现只要系统有足够的时间和能量驱动网络收敛到真值,即使随后网络达到了同步状态,网络辨识的结果也不会再发生改变,在这种情况下网络同步并没有阻碍结构的辨识。
[Abstract]:With the rapid development of the times and the data of information technology is coming, people enjoy the network to bring more convenience, more and more data and information in the transmission process is also faced with monitoring, theft, tampering and other threats. The issue of information security has become a major problem in this age affect the national security and social development, secure communication the existing methods are facing more and more challenges, the new research method of secure communication to protect the security of information and data communication become the communications, computer, information and other related fields in the frontier and hot direction.
Chaotic signal has the sensitivity to initial conditions, non periodic, noise, continuous broadband spectrum property, which has a high degree of unpredictability and natural concealment, especially as the carrier of secure communication. At the beginning of 1990s, the idea of chaotic secure communication was developed, will become hot issues in the field of information security, domestic and foreign scientists and researchers from the information encryption algorithm, robustness, different angles of chaotic devices done in-depth research on chaotic secure communication model and its related technology. Now in the era of big data, the research direction of chaotic secure communication towards the digital, multi user high performance broadband, wireless, and other aspects of development.
Typically, chaotic system, chaotic secure communication model exists in part or all of the parameters of unknown parameters is unknown, this is unknown although will make the system more difficult to operate and control methods, it will be for information modulation by using these parameters to bring more security, improve the security of communication in order to process. Better operation of this kind of system and the realization of information encryption, can be adaptive parameter identification of nonlinear control theory in the introduction of chaotic systems with unknown parameters, through continuous measurement of the system state or output, compared with the expected target, according to the structure and the parameter of this difference adjustment system to realize the control function, until the cycle to achieve the ultimate goal.
In this paper, the adaptive parameter identification technology of chaotic system and its application in secure communication are studied. The main innovation work is summarized as follows:
(1) this paper proposes an adaptive parameter identification in new condition - Gram matrix condition, put forward the conditions established between persistent excitation condition and linear independence condition of the bridge. The persistent excitation condition or linear independence condition are two basic conditions for wide application of adaptive parameter identification, but the relationship between them has been in suspense. The calculation method is different from the two conditions, convergence of Gram matrix condition proposed by calculating the rank of matrix to determine the unknown parameters, in some cases, can reduce low computation calculation of persistent excitation condition or linear independence condition, more practical and efficient.
(2) this paper pointed out that in the design of identification error when many articles in the article. The definition of adaptive parameter identification in traditional linear independent conditions for the lack of the time limit for a time-varying function, not by ordinary linear independent definition and judge the correlation. This thesis presents the persistent excitation condition the definition, linear independence condition or Gram matrix condition in long time and limited time conditions, the influence of the synchronization process of chaotic system and the criteria for the duration of the identification results. Besides, this paper puts forward the method of parameter identification under the presence of noise, and measures to reduce noise. To determine the conditions and the method proposed in this part can be used in chaotic secure communication scheme.
(3) this paper presents a model of adaptive chaotic communication based on parameter identification, the one-dimensional Scalar Chaotic signal to load multiple information, realize the chaotic secure communication transmission of information. In this communication model, the original information is first parameter Scalar Chaotic System of direct modulation in the transmitter with multiple delay in the demodulation of loading in chaos signal information at the receiver through the method of adaptive parameter identification. This is carried by the one-dimensional chaotic communication model of chaotic system structure of the information carrying amount of information system and much more than the traditional chaotic communication model is more simple. Then, based on the orthogonal basis of Walsh function, the the orthogonal modulation multiple original information and Walsh sequence, chaotic system parameters and loading at the transmitting end, greatly improve the system to carry the information The chaotic secure communication scheme proposed in this paper has potential applications in the field of chaotic laser communication.
(4) this paper presents a method of using chaotic synchronization to realize the transient process of complex dynamic network system structure identification. The research of chaotic secure communication scheme composed of two chaotic systems, the chaotic communication transmission point-to-point. Multiple chaotic communication system mutual communication constitute the structure of complex dynamical networks the study of complex dynamic network. In recent years the article generally believe that the synchronization phenomenon within the network topology identification is completed hinder. This paper explores the transient process of network synchronization, using the method of parameter identification of the adaptive research of complex network topology the internal links between the nodes, the system found that as long as there is enough time and energy to drive network converge to the true value, then even if the network has reached the synchronization state, network identification results will not change, in this case the network synchronization and There is no hindrance to the identification of the structure.

【学位授予单位】：北京邮电大学
【学位级别】：博士
【学位授予年份】：2014
【分类号】：TN918;O415.5

【参考文献】