【作者】 李赞；

【导师】 常义林；

【作者基本信息】 西安电子科技大学 ， 通信与信息系统， 2006， 博士

【摘要】 近年来，随着灾害预警、军事通信、网络安全、信息隐匿、抗干扰系统等各种通信应用领域对于高安全性、高可靠性的通信新要求，隐蔽突发传输以可靠、及时、准确、安全的系统性能及其在保障国民经济和人民生命财产安全、军事战略应急、最低限度通信保障等方面的巨大应用意义和价值跃然成为国内外研究的热点。在国内，作为刚刚起步的崭新研究领域，隐蔽突发传输的许多关键技术问题急待解决。 隐蔽突发传输在广域意义上涵盖了广阔的研究范畴，随着通信新应用、新需求、新概念的不断产生，其定义不仅仅局限于传统的传输信息自身的隐蔽上，还应当体现在传输方式的隐蔽、传输机制的隐蔽、传输信道的隐蔽、信号频谱的隐蔽等方面，从而满足实际的通信应用和需求。本论文在给出了广域意义上隐蔽突发传输新定义的基础上，以自适应信号处理和频率认知为切入点，对其理论和关键技术进行了深入的研究，论文主要内容与成果如下： 1．以典型的隐蔽突发传输信道——流星余迹信道为例，以南极中山站～昭和站约1400公里的流星余迹信道极区测试数据和陕西省西安市～河北省廊坊地区霸县约1000公里的远距离流星余迹通信试验测试数据为基础，设计基于突发信道特性的传输帧结构，提出了基于实时信道估计的自适应变速率传输机制。在此基础上，将突发传输信号的接收过程划分为两个阶段：信道参数捕获阶段、信道跟踪与均衡阶段。分别构建了两个阶段的系统传输模型，推导了各阶段模型下的系统信道估计性能。证明在理想情况下均有P_opt（Z）=H（Z），即各阶段均能够准确的捕获信道参数。 2．在建立的信号接收模型的基础上，研究快速信道估计算法和均衡算法；根据信道跟踪与均衡阶段信号传输和接收模型，提出了基于快速复信道参数捕获的系数映射判决反馈均衡算法，推导了实时复信道参数与判决反馈均衡器（DFE：Decision Feedback Equalization）抽头系数之间的直接映射关系，从而实现了基于实时信道估计的有效数据均衡。仿真结果和性能分析表明，该算法能够避免传统判决反馈均衡器自适应训练过程的系数不收敛和跟踪延迟问题，适合于突发通信的自适应变速率通信机制，并具有较好的抗噪声能力及较低的计算复杂度。 3．针对具有理论最佳性能的联合信道跟踪与最大似然检测理论（PSP：Per-Survivor Processing）存在的计算复杂度较大的问题，以及国内外现有各类准PSP算法的局限，首先从静态处理的角度出发，提出了基于降阶的联合信道跟踪与最大似然检测算法（D-PSP：Dimension-down PSP），从前后码元的相关性程度出发，对PSP联合检测进行降N阶处理。根据信道特性选择不同的降阶维数N，D-PSP算法在保证数据正确接收率的条件下，可以不同程度的有效减小PSP算法的计算时间和存储空间，使隐蔽突发传输自适应编码调制方式数据的最大似然接收成为可能。其次，从动态处理的思路出发，针对典型的流星余迹信道随时间呈指数规律衰减的特性，提出了基于动态自适应状态分配的联合信道跟踪与最大似然检测理论（ASP：Adaptive State Reduction of PSP）。在系统参数估计的基础上，推导出适合于流星余迹通信的实时数据接收状态的删移门限Th_k^max，在数据接收过程中动态的控制联合信道跟踪与最大似然接收的状态数，有效减小PSP算法的运算复杂度和存储空间，实现流星余迹通信的自适应可靠数据检测与接收。 4．针对国内外现有跳频序列族（Prequency Hopping Sequences Family）产生理论存在的局限性，首次将密码学的加密思想引入到隐蔽突发传输的跳频地址序列族设计之中，提出了基于分组密码的FH（Prequency Hopping）序列族产生方法，使生成的序列具有优异的综合性能指标。在此基础上，对于跳频序列的安全性、随机性（独立性）、均匀性、长周期性、频率撞击特性等各项系统性能指标进行了检验。同时，建立了跳频码分多址（FHMA：Frequency Hopping Multiple-access）组网接入模型，对于跳频序列的跳频码分多址接入性能进行了理论推导，得出了跳频组网的系统误分组率、吞吐量和归一化吞吐率的表达式和结论，并对现有典型的跳频序列族的FHMA组网性能进行了检验和分析，得出了重要结论。 8．针对目前国内外普遍采用的LC（Linear Complexity）、LCP（Linear Complexity Profile）和k-error LC（k-error Linear Complexity）序列复杂度检测理论存在的缺陷，将物理学和信息论中信息熵的概念引入到序列的复杂度检验中，提出了基于近似熵（ApEn：Approximate Entroy）的FH/SS（Frequnecy Hopping/Spread Spectrum）序列复杂度检验理论。给出了用ApEn作为度量混沌运动系统和隐蔽突发传输离散跳频地址序列的复杂度准则，并针对FH/SS序列的复杂度测度进行了新定义，证明了ApEn的两个重要结论。然后，利用随机统计学原理，分析了ApEn计算中的参数的选取问题，并以Logistic映射和耦合映象格子映射等典型FH/SS地址序列族为例，验证了理论和分析的正确性。同时，提出了基于信息熵的无线信道频谱预测和认知方向的进一步应用。更多还原

【Abstract】 Recently, with exponentially growing demands for high security and reliable data transmission in applications of disaster early warning, military communication, network security, emergency communications, information hiding and interference cancellation, hidden data transmission has attracted a wide range of attention for its great importance, urgent needs and encouraging prospects. As a new field of research in our country, hidden data transmission can provide high performance of communication with reliability, accuracy and security, whereas most of its key technologies are in great need of deep study.With the rapid developments of new concepts and applications, hidden data transmission embraces much more fields of science. To satisfy the critical requirements of practical communications, hidden data transmission is not restricted to the traditional information hiding and should be extended to the hiding of communication manner, transmission mechanism, wireless channel and frequency spectrum, etc. Given the new definition of hidden data transmission in broad sense, the research work in this dissertation carries a rather deep study on the key technologies of adaptive signal processing and frequency recognition for hidden data transmission, and the major achievements and results are outlined as follows.1. Taking the typical hidden communications—meteor burst communication (MBC) for instance, the structure of its transmission data frame is constructed and the mechanism of adaptive data transmission based on channel estimation is proposed. Adopting the collected data of long distance field try from Zhongshan Station to Syowa Station in Antarctic and from Xi’an city of Shaanxi province to Langfan city of HeBei province in China, the characteristic of meteor channel is analyzed which provide theoretical basis for the realization of MBC system. On this basis, a model of signal processing based on the structure of data frame is suggested. There are two distinct modes of operation for signal processing: acquisition and tracking. And the performance of channel estimation for each mode is derived.2. According to the analysis of hidden data transmission mechanism, a receiver based on coefficients-mapped decision feedback equalization (DFE) via channel acquisition is suggested in accordance with the data frame structure. With the channel impulse response being obtained by fast correlation, the relation between channelparameters and equalizer coefficients is derived and the DFE coefficients are directly calculated from the estimated channel parameters, which shows more significantly robust against channel variation and noise than a DFE employing adaptive coefficient tracking does, and therefore makes data receiving stable for adaptive modulation and coding (AMC) of hidden communication system. Computer simulation results are included to support our developments.3. For joint data and channel estimation, per-survivor processing (PSP) provides superior performance and robustness whereas the giant computational complexity limits its application in adaptive hidden data transmission. Here, a suboptimal method called D-PSP is proposed to save the computational time and memory size, which agrees with the slow-fading characteristic of burst channel and makes the MLSD possible for adaptive data transmission. In the meanwhile, an adaptive state reduction of PSP (ASRP) algorithm is presented with only a few states in the trellis diagram being employed, while these states are chosen by the time-varying threshold according to the exponential decay of meteor channel. It is shown that, ASRP can make good tradeoff between performance and computational complexity dynamically, and provides reliable data transmission for adaptive modulation and coding of MBC system.4. In view of the inherent weaknesses of FH sequence families available in various fields, a novel family of frequency hopping sequences based on iterated block cipher is proposed for frequency-hopping multiple-access (FHMA) burst communications. The design offers a class of nonlinear FH codes with high security, large linear span and a uniform spread over the entire frequency bandwidth. Moreover, FH sequences among the family are independent from each other and they perform as well as random patterns in terms of multiple access interference in anti-jamming applications. With the performance of packet error and throughput for FHMA network being derived, the simulation results also show that the proposed sequences have multiple access performance identical to that derived on the basis of random sequence model. Efficiently implemented in field-programmable-gate-arrays (FPGA), the generator prototype of the proposed sequence has been realized and incorporated into fast FH radios.5. High complexity of spread-spectrum (SS)/frequency-hopping (FH) sequence is of great importance for high-security-required multiple-access communications. With the growing developments of the construction of SS/FH sequence in much wider fields, the well-known complexity measures—the linear complexity (LC), the linear complexity profile (LCP) and the k-error linear complexity (k-error LC)—are widelyused but not sufficient to evaluate the complexities of the sequences available, for instance, the cryptographical sequence and the chaotic sequence families. In this paper, a novel complexity metric to evaluate the unpredictability of SS/FH sequence is proposed based on the concept of the approximate entropy (ApEn), which is defined as the maximal randomness of sequences with arbitrary length. The bounds of the ApEn are derived and it is shown that, by appropriately choosing its parameters, the ApEn can effectively discern the changing complexities of SS/FH sequences from a relatively small amount of samples, which is superior to its candidates and can be extended to frequency prediction for cognitive radios. Simulation results and analysis are also included to support our developments.更多还原