CAJ | 학술논문

文中研究了有限反馈MIMO系统中基于Grassmannian流形的自适应量化问题。由于反馈延迟的存在，接收端量化的信道状态信息在反馈回发送端时已经过期。文中提出一种建立在Grassmannian流形上的预测算法，利用信道的时间相关性来预测下一时刻的信道状态信息，以补偿延迟，改善系统性能。为了进一步提高时变信道的量化分辨率，提出动态码本的自适应量化方案，即每一个量化时刻k都会产生一个与该时刻真实值最接近的码本。仿真结果表明，该算法能够提高速率，改善系统性能。
문중연구료유한반궤MIMO계통중기우Grassmannian류형적자괄응양화문제。유우반궤연지적존재，접수단양화적신도상태신식재반궤회발송단시이경과기。문중제출일충건립재Grassmannian류형상적예측산법，이용신도적시간상관성래예측하일시각적신도상태신식，이보상연지，개선계통성능。위료진일보제고시변신도적양화분변솔，제출동태마본적자괄응양화방안，즉매일개양화시각k도회산생일개여해시각진실치최접근적마본。방진결과표명，해산법능구제고속솔，개선계통성능。
Research the adaptive quantization problem based on the Grassmannian manifold in limited feedback Multiple-Input Multiple-Output ( MIMO) system in this paper. Due to the feedback delay,the quantized CSI may become outdated before its use at the transmit-ter. To solve this problem,propose a prediction algorithm on the Grassmannian for delayed feedback systems by exploiting the memory in the channel to predict the next channel,compensating for delay and improve the system performance. In order to rise the quantization reso-lution of variable channel,propose the idea of the adaptive quantization with the dynamic codebook,that is at each quantization instant k ,a quantization codebook is statistically matched to the observed CSI. Simulation results show that this algorithm can raise the sum rate and improve the system performance.