常用数字信号调制方式识别研究-通信与信息系统专业论文.docxVIP

I I 摘 要 摘 要 数字通信信号调制方式识别历来是各国研究的热点，在军用和民用通信领域都有着 广泛的应用前景。近年来研究者将神经网络、支持向量机、核方法、模糊等智能模式识 别方法与调制识别相结合，提出了很多新算法。本文通过对已有的调制特征进行选择和 提取，并构造新的适用特征量，提出了基于最小距离和模糊匹配相结合的数字信号调制 方式统计模式识别算法。 本算法不仅仅用单一的特征参数阈值对信号进行简单的分类，而是进行多参数联合 识别。在类间识别时，利用最小距离分类算法，在信噪比大于 8db 的情况下，能有效区 分 ASK、PSK、FSK、QAM 信号类型。在类内识别时，利用模糊匹配分类算法，结合 阈值识别 QAM 信号，在信噪比大于 8db 的情况下，能有效区分 ASK、PSK、FSK、QAM 信号的进制。针对低信噪比条件下 FSK、PSK 信号较难识别的问题，利用二阶矩、四阶 矩估计算法对受噪声污染的 FSK、PSK 信号信噪比进行了估计，并研究了信噪比对特征 参数的影响，从而为系统实时修改特征参数值提供了依据。通过大量的计算机仿真实验， 验证了该算法在信噪比大于 8db 条件下对 ASK、PSK、FSK 和 QAM 信号的分类正确识 别率都接近 98%，相比已有算法，进一步提高了系统的分类性能。 关键词 数字调制识别 特征选择与提取 最小距离 模糊匹配 II II Abstract Abstract Modulation recognition of Digital Communication Signal has always been a research focus for countries, which has a wide range of applications in military and civilian communication fields. In recent years, the researchers combined with intelligent pattern recognition methods such as neural networks, support vector machines, kernel methods, fuzzy, etc. proposed a lot of modulation recognition methods. By selecting and extracting some original modulation characteristics, and constructing a new applicable feature, this paper put forward a new Statistics Pattern Recognition Algorithm which based on minimum distance and fuzzy matching. The algorithm does not just use a single threshold of parameters to simply classify signals, but rather a multi-parameter joint identification. To identify the inter-class, using a minimum distance classification algorithm, can effectively distinguish ASK, PSK, FSK, QAM signals when the SNR is greater than 8db. To identify the type, using fuzzy matching algorithm and combined with the threshold to identify QAM signal, can effectively distinguish the decimal of ASK, PSK, FSK, QAM signals when the SNR is greater than 8db. To solve the problem that FSK, PSK signal is difficult to identify in low SNR conditions, using second moment, fourth moment algorithm to estimate the SNR of noise contaminated FSK, PSK signal, and studied the SNR impact to