摘要
针对多模辅助算法均方误差高的缺点,提出了两种基于星座图改进的盲均衡算法.首先利用16QAM星座图信息,增加多模辅助算法代价函数中信号点的均衡数目,得到改进的多模辅助算法,与多模辅助算法相比,改进后的算法稳态均方误差更低,而且收敛速度略有提高;其次将改进的多模辅助算法和判决引导算法相结合,根据均方误差调整两种算法的比例,得到的混合算法进一步降低了稳态均方误差.水声信道仿真结果表明:提出的两种算法都可以有效地降低均方误差,适用于对精度要求较高的场合.
Two improved blind equalization algorithms based on constellation chart are put forward, aiming to solve the high mean square error in MMA-Assisted Modulus Algorithm. Firstly, the 16QAM constellation chart information is used to increase the signal point number of cost function of MMA-Assisted Modulus Algorithm, thus get the improved blind equalization algorithm, which reduces the steady-state mean square error and slightly improves the con- vergence speed of equalizer. Secondly, integrate the improved blind equalization algorithm and decision direction algorithm, adjust the proportion of the two algorithm according to the mean square error, and get a hybrid algorithm, which can further reduce the steady-state mean square error. The acoustic channel simulation results show that the two improved blind equalization algorithms have smaller mean square error, and are suitable for applications with high accuracy requirement.
出处
《南京信息工程大学学报（自然科学版）》
CAS
2012年第5期447-451,共5页
Journal of Nanjing University of Information Science & Technology（Natural Science Edition）
基金
江苏省自然科学基金(BK2009410)
关键词
水声信道
盲均衡
正交幅度调制
星座图
acoustic channel
blind equalization
quadrature amplitude modulation
constellation chart
