信号分割修正聚类的星座图恢复算法

Constellation Rebuilding Algorithm Using Signal Segmentation and Correction Clustering

通过分析载波估计频偏、信号长度对星座图恢复的影响,采用将长信号分割为短信号的方法减小频偏对星座图恢复的干扰,并通过旋转叠加修正的方法解决了由信号长度减小导致码元失去历遍性而带来的星座图恢复不完整问题,最后根据信噪比大小选取不同聚类算法进行聚类并再次修正,实现了星座图对称的幅相调制信号的星座图完整恢复。多次仿真表明,该算法对频偏估计精度的要求可以降低1~2个数量级,减小了参数估计的工作量,使因参数估计精度不高而带来的信号后续分析困难问题得到了良好解决。

By analyzing the impact of carrier frequency estimation offset and signal length on rebuil- ding of the constellation diagram, a method of dividing long signal into several short signals is adopted to reduce the interference of frequency offset to constellation diagram recovery, and the problem that the chip without ergodic is solved by using the correction method of whirling-superposition. Finally, accord- ing to the value of SNR, different clustering methods are used to correct it again. The above measures a- chieve constellation rebuilding of amplitude-phase modulation signals, whose constellations are symmet- ric. Simulation results show that the algorithm for frequency offset estimation accuracy requirements can be reduced by one to two orders of magnitude and it reduces the parameter estimation work quantity and it can solve the problem that the Signal analysis and processing cannot continue because of low Parameter estimation accuracy.