摘要
阐述了恒模算法(CMA)、修正恒模算法(MCMA)和判决引导(DD)算法的基本原理。针对CMA和MCMA收敛速度慢,固定步长条件下收敛速度和剩余误差之间存在矛盾的缺陷,在分析CMA误差特性的基础上,利用DD算法误差函数的模值和判决器输出构造新的误差函数,提出了一种新的变步长MCMA。新算法在加快收敛速度的同时保持小的剩余误差。仿真结果表明,新算法比CMA和MCMA收敛速度快,均衡输出剩余码间干扰(ISI)小而且能克服相位偏移,具有很好的实用性。
The basic theories of constant modulus algorithm (CMA), modified CMA (MCMA) and direct decision if)D) algorithm are expounded firstly. Aiming at the disadvantages of CMA and MCMA such as low convergence speed, conflict between convergence speed and residual error at constant step size, a new variable-step MCMA is proposed, which uses the modulus of DD algorithm's error function and decision device's output to construct a new error function. The new algorithm can speed up the convergence and keep less residual error simultaneously. The simulation results show that the new algorithm has faster convergence, smaller intersymbol interference (ISI) than CMA and MCMA and can eliminate the phase offset. So it is of practical inportance.
出处
《计算机与网络》
2012年第15期61-63,共3页
Computer & Network
关键词
修正恒模算法
收敛速度
剩余误差代价函数
变步长
modified constant modulus algorithm (MCMA)
convergence speed
residual error
cost function
variable step size