Simplified Approach to Optimized Iterative Clipping and Filtering for PAPR Reduction of OFDM Signals
Simplified Approach to Optimized Iterative Clipping and Filtering for PAPR Reduction of OFDM Signals
复制标题
DOI：
10.1109/tcomm.2013.021913.110867
复制
发表时间：
2013-02
影响因子：
8.3
通讯作者：
Xiaodong Zhu;Wensheng Pan;Hong Li;Youxi Tang
中科院分区：
文献类型：
--
作者：
Xiaodong Zhu;Wensheng Pan;Hong Li;Youxi Tang
迭代限幅滤波（ICF）是一种众所周知的降低正交频分复用（OFDM）信号峰均功率比（PAPR）的技术。最近，王和罗对限幅信号进行了研究，并提出了一种改进算法，称为优化迭代限幅滤波（OICF）。这是一种最优算法，因为它能够以最小的带内失真和少得多的迭代次数实现所需的PAPR降低。然而，OICF需要解决一个复杂度为O(N³)的凸优化问题，其中N表示子载波的数量。在本文中，我们不分析限幅信号，而是研究限幅噪声，并提出一种简化的OICF算法。在新算法中，通过一些简单的代数运算来近似解决凸优化问题，计算复杂度降低到O(N)。仿真结果表明，经过三次迭代，原始的OICF算法能够实现期望的PAPR，而简化算法表现出几乎相同的性能：对于一个具有128个子载波且采用正交相移键控（QPSK）调制的OFDM系统，在限幅概率为10⁻⁴时，两种算法之间的PAPR降低性能差异为5×10⁻³dB，在误码率为10⁻⁷时，误码率性能差异为6×10⁻³dB。
Iterative clipping and filtering (ICF) is a well-known technique to reduce the peak-to-average power ratio (PAPR) of orthogonal frequency division multiplexing (OFDM) signals. Recently, Wang and Luo investigated the clipped signal and proposed a modified algorithm called optimized ICF (OICF). This is an optimal algorithm since it can achieve the required PAPR reduction with minimum in-band distortion and far fewer iterations. However, OICF needs to solve a convex optimization problem with O(N3) complexity, where N represents the number of subcarriers. In this paper, instead of analyzing the clipped signal, we study the clipping noise and propose a simplified OICF algorithm. In the new algorithm, solving the convex optimization problem is approximated by some simple algebraic operations and the computational complexity reduces to O(N). Simulation results show that after three iterations, the original OICF algorithm can achieve the desired PAPR while the simplified one exhibits almost the same performance: for a 128-subcarrier and quadrature phase shift keying (QPSK) modulated OFDM system, the PAPR-reduction performance difference between the two algorithms are 5×10-3dB at a 10-4 clipping probability and the bit-error-rate performance difference is 6×10-3 dB at a 10-7 error probability.