利用多级线性预测改善有限反馈系统反馈量

Improving amount of feedback in limited feedback systems with multi-level linear prediction

以先进的长期演进技术(LTE-Advanced)为背景,针对多输入多输出(MIMO)系统中基于码本的有限反馈技术进行了研究,提出了一种基于多级线性预测的有限反馈新方法。利用信道的时间相关性,对信道值进行多级线性预测,以预测误差为基础,通过最小化均方误差来设计量化码本,将码本中对应的码字序号用于系统反馈。仿真结果表明:采用多级线性预测能有效地降低系统预测误差,等效地,减小了系统的反馈开销,最大降低幅度能达到15%。

In this paper, the limited feedback techniques based on the codebook for Multiple-Input Multiple-Output （MIMO） system in the Long Term Evolution-Advanced （LTE-Advanced） applications were investigated. A new limited feedback technique based on muhi-level linear prediction was proposed. This method utilized the time correlation of the channel to predict the value of the channel based on multi-level linear prediction. By minimizing the mean square error to design a quantization code book in the basis of the prediction error, code number in the code book was used for system feedback. The simulation results show that muhi-level linear prediction can effectively reduce the system prediction error, equivalently, reducing feedback overhead of the system and the maximum reduction rate of 15% to the amount of system feedback that can be reached.