脉冲雷达信号分选识别算法研究

发布时间：2018-01-06 00:29

本文关键词：脉冲雷达信号分选识别算法研究　出处：《电子科技大学》2016年硕士论文　论文类型：学位论文

【摘要】：脉冲雷达信号的分选识别,作为电子侦察中最为重要的一环,是雷达侦察的基础与核心。在当今复杂的电磁空间环境中,不但信号密度大,而且信号脉内调制和脉间调制也是愈发复杂,这使得雷达侦察的难度也越来越大。现代电子情报系统(Electronic Intelligence,ELINT)、电子支援系统(Electronic Support Measure,ESM)和雷达威胁警报系统(Radar Warning Receiver,RWR)都要求具备准确分析、快速处理各种雷达信号的能力,这也对雷达信号的分选识别的准确性与实时性有很高的要求。本文针对脉冲雷达信号的分选识别与多核数字信号处理器(Digital Signal Processor,DSP)上的硬件实现做了以下工作:1.介绍了雷达信号分选识别的脉内参数及脉间参数,分析了各参数在信号分选识别中的地位。研究了现代雷达信号分选中较常用的几种雷达分选算法,进行了仿真实验,比较了多种雷达信号分选算法的优劣及适用范围。2.针对现有的分选算法对复杂调制脉冲重复间隔(Pulse Repeatition Interval,PRI)分选效果不理想的情况,研究了一种可用于工程实现的信号分选算法;利用脉冲到达角(Direction of Arrival,DOA)、脉冲载频(Carrier Frequency,CF)、脉冲宽度(Pulse Width,PW)等参数进行脉冲相似度分析,研究了基于脉冲相似度的改进序列差分直方图(Sequential Difference Histogram,SDIF)算法。3.针对目前雷达常用的四种PRI调制样式:固定PRI,抖动PRI,滑变PRI以及参差PRI,研究了一种快速稳健的PRI调制识别方法。该方法通过对PRI序列提取四种特征参数,可在脉冲丢失率小于20%的情况下达到90%以上的正确识别概率。且该识别算法运算量较小,可在工程中快速实现。4.本文以一个ELINT原型样机的工程实例为背景,介绍了雷达信号分选识别算法实现的硬件平台。基于TMS320C6678硬件平台研究了脉冲雷达信号分选识别算法的硬件实现。利用TMS320C6678芯片的多核结构相对传统单核DSP的优势,研究了一种雷达信号侦察算法多核分配结构,将分选识别算法在DSP上进行了硬件实现。最后给出了系统联试的测试结果,证明了信号分选识别算法的有效性和多核可实现性。
[Abstract]:As the most important part of electronic reconnaissance, the sorting and recognition of pulse radar signal is the basis and core of radar reconnaissance. In the complex electromagnetic space environment, the signal density is not only high. Moreover, intra-pulse modulation and inter-pulse modulation are becoming more and more complex, which makes radar reconnaissance more and more difficult. Modern electronic information system electronic Intelligence. Electronic Support Measure. Both ESMand Radar Warning receiver (RWR) require the ability of accurate analysis and fast processing of various radar signals. This also requires the accuracy and real time of radar signal sorting and recognition. This paper aims at the sorting and recognition of pulse radar signal and multi-core digital signal processor (. Digital Signal Processor. The hardware implementation of DSP has done the following work: 1. The intra-pulse parameters and inter-pulse parameters of radar signal sorting and identification are introduced. This paper analyzes the status of each parameter in signal sorting and recognition, studies several radar sorting algorithms commonly used in modern radar signal sorting, and carries out simulation experiments. The advantages and disadvantages of various radar signal sorting algorithms and their application range are compared. 2. Aiming at the existing sorting algorithms for complex modulation pulse repetition interval (. Pulse Repeatition Interval. In the case of poor sorting effect, a signal sorting algorithm which can be used in engineering is studied. Pulse arrival Angle Direction of Arrivalo load, Pulse Carrier Frequency (CF). Pulse width pulse width and other parameters are analyzed by pulse similarity analysis. The improved sequence differential Difference Histogram based on pulse similarity is studied. SDIF.3. aiming at four kinds of PRI modulation styles commonly used in radar at present: fixed PRI, jitter pri, sliding PRI and staggered PRI. A fast robust PRI modulation recognition method is studied, which extracts four characteristic parameters from PRI sequences. When the pulse loss rate is less than 20%, the correct recognition probability can be achieved above 90%, and the computation of the recognition algorithm is small. It can be quickly realized in engineering. 4. The background of this paper is an engineering example of a prototype of ELINT. This paper introduces the hardware platform of radar signal sorting recognition algorithm. Based on TMS320C6678 hardware platform, the hardware implementation of pulse radar signal sorting recognition algorithm is studied. TMS320C66 is used to realize the algorithm. The multi-core structure of the 78 chip is superior to that of the traditional single-core DSP. In this paper, a multi-core allocation structure of radar signal reconnaissance algorithm is studied, and the sorting and recognition algorithm is implemented on DSP. Finally, the test results of the system are given. It is proved that the algorithm of signal sorting and recognition is effective and multi-core realizable.
【学位授予单位】：电子科技大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：TN957.51

【参考文献】