欢迎您！东篱公司
退出
-
申报数据库
-
立项数据库
-
成果数据库
-
项目获奖数据库
中文摘要：
为了改善扩展的二元相移键控（EBPSK）系统在低信噪比下的误码率性能,针对解调端冲击滤波器输出信号的特点,引入支持向量机（SVM）分类方法.在滤波器输出的中频信号中选取少量采样点进行判决,仿真显示,可以得到较高的信噪比增益.误码率在10-4时,比积分判决方法获得1.8 dB的信噪比增益.不同核函数产生不同的支持向量机算法,进而对线性和径向基核函数作了分析,同时,对不同的特征点提取以及不同的训练码元个数对判决结果的影响作了较为详细的分析.通过仿真发现,用少量特征点和训练码元便可以获得较好的性能,因此,在EBPSK系统中采用SVM分类判决法降低误码率是一种较有效的方式.
英文摘要：
To improve the bit error rate（BER） performance of the extended binary phase shift keying（EBPSK） demodulation under low signal to noise ratio（SNR） and according to the signal characteristics of impact filter output,the support vector machine（SVM） classification is introduced.A few signal samples are selected from the intermediate-frequency output of the filter to make the judgment.Simulation reveals that a higher SNR gain is obtained by adopting the SVM classification.Compared with integral decision 1.8dB higher SNR gain can be obtained when BER is 10-4.Different kernel functions have different support vector machine algorithm,and then the radial basis function kernel and linear functions are analyzed.The effect of different characteristic point Abstraction and different numbers of training elements on discrimination results are analyzed in detail.It is found from the simulation that using only a few characteristic points and training elements can achieve better results.Therefore,it is an effective way adopting SVM classification in EBPSK system to reduce BER.
同期刊论文项目
同项目期刊论文
位置：