当发射端信道状态信息可以通过反馈或时分双工的互逆原理得到时,奇异值分解法可将MIMO信道分解为若干并行子信道以获得最大的吞吐量。然而,由于奇异值分解得到的子信道的信噪比差别很大,采用自适应编码调制技术时,该方案需要根据不同子信道的信噪比配置不同的编码、调制参数。为此,研究基于几何均值分解的MIMO复用系统下行链路误码率性能,导出其误码率表达式。研究结果表明,采用几何均值分解得到的各个子信道的误码率相同,因而便于编码、调制技术的采用。
To maximize the channel throughput in MIMO systems, the channel must be diagonalized via the singular value decomposition ( SVD), when channel state information can be obtained at the transmitter via feedback or the reciprocal principle with using time division duplex (TDD). However, due to the very different SNRs of the subchannels, this scheme requires different parameters for adaptive modulation and coding (AMC) to match the subchannel BERs. To study the BER performance of the MIMO multiplexing system with geometric mean decomposition (GMD) in the downlink, exact BER expression is derived. Results show that the BER performances are the same for the different subchannels diagonalized via the GMD, which bring about much convenience in coding/decoding and modulation/demodulation processes.