On the Diversity Behavior for Fixed-Gain Amplify-and-Forward Relaying Over Frequency-Selective Channels Based on Linearly Precoded OFDM
On the Diversity Behavior for Fixed-Gain Amplify-and-Forward Relaying Over Frequency-Selective Channels Based on Linearly Precoded OFDM
复制标题
基于线性预编码正交频分复用的频率选择信道固定增益放大前转分集行为研究
DOI：
10.1109/tvt.2012.2191426
复制
发表时间：
2012
影响因子：
6.8
通讯作者：
Jintao Sun
中科院分区：
文献类型：
--
作者：
Qingchuan Zhang;F. Shu;Jintao Sun
最近的研究表明，固定增益放大转发（FG-AF）中继之间的合作的分集增益是不是一个简单的函数的信噪比（SNR）的功率，其中可能涉及的SNR的对数的一个因素。用ρ<sup>-y</sup>（log ρ）<sup>x</sup>的分集增益函数（DGF）来描述这种分集性能更为合适，其中ρ表示信噪比，<i>x</i>和<i>y</i>为非负整数。在这封信中，我们考虑使用FG-AF中继频率选择性信道的基础上的正交频分复用（OFDM）传输采用线性预编码（LP），并提出全面的分析所产生的DGF。假设从源到中继的信道的多径分量的数量是<sub>L1</sub>，从中继到目的地的多径分量的数量是<sub>L2</sub>。<i></i><i></i>理论分析表明，当<i>L1</i><sub><</sub>$L2<sub>时</sub>，只要预编码组大小<i>J</i>≥<i>L</i> + 1，最优DGF就是ρ<sup>-L</sup>，<i>L</i>= min{<sub>L1</sub>，<sub>L2</sub>}，而预编码组大小<i>J</i>≤<i>L时</i>，最优DGF只有ρ<sup>-J</sup>（log ρ）<sup>J</sup>.<i></i><i></i><i></i>在<sub>L1</sub>=<sub>L2</sub>的情况下，我们证明了最佳可达DGF是ρ<sup>-Llog</sup> ρ，这可以通过选择<i>J</i>≥<i>2L</i>- 1来保证.<i></i><i></i>我们还证明，这些结果可以很容易地扩展到多继电器的情况下。
Recent research has shown that the diversity gain of cooperation among fixed-gain amplify-and-forward (FG-AF) relays is not a simple function of power of signal-to-noise ratio (SNR), where a factor of the logarithm of SNR might be involved. It is more suitable to use the diversity gain function (DGF) of ρ<sup>-y</sup>(log ρ)<sup>x</sup>, with ρ denoting the SNR and nonnegative integers <i>x</i> and <i>y</i>, to describe such diversity performance. In this letter, we consider the use of FG-AF relaying over frequency-selective channel based on the orthogonal frequency-division multiplexing (OFDM) transmission employing linear precoding (LP) and present comprehensive analysis of the resulting DGF. Suppose that the number of multipath components for the channel from source to relay is <i>L</i><sub>1</sub> and that from relay to destination is <i>L</i><sub>2</sub>. Theoretical analysis shows that, if <i>L</i><sub>1</sub> ≠ <i>L</i><sub>2</sub>, the optimal DGF is precisely ρ<sup>-L</sup>, with <i>L</i> = min{<i>L</i><sub>1</sub>, <i>L</i><sub>2</sub>}, which can be achieved as long as the precoding group size <i>J</i> ≥ <i>L</i> + 1, whereas only a DGF of ρ<sup>-J</sup>(log ρ)<sup>J</sup> is obtained by precoding with size <i>J</i> ≤ <i>L</i>. In the case of <i>L</i><sub>1</sub> = <i>L</i><sub>2</sub>, we show that the best achievable DGF is ρ<sup>-L</sup>log ρ, which can be guaranteed by choosing <i>J</i> ≥ 2<i>L</i> - 1. We also demonstrate that these results can be easily extended to the multirelay case.