论文标题
MU-MIMO通道的对角线排列实现的能力
Capacity Achieving by Diagonal Permutation for MU-MIMO channels
论文作者
论文摘要
肮脏的纸质编码（DPC）被认为是达到高斯多输入多输出（MIMO）广播通道（BC）的最佳预编码。但是，要找到最佳的预编码顺序，它需要重复N！对于n个用户来说，有n个时间！可能的预编码订单。这种极高的复杂性限制了其在现代无线网络中的实际使用。在本文中，我们在2-D（空间）情况下显示了DPC的等效性以及最近提出的高阶Mercer定理（HOGMT）预编码[1] [2]，这为DPC提供了替代实现。此外，我们表明，在多用户通道上置换时，所提出的实现方法对置换操作员是线性的。因此，我们提出了一种低复杂性算法，该算法优化了通过波束形成的DPC的预编码顺序，从而消除了每个预编码顺序的重复计算DPC。仿真表明，我们的方法可以达到与常规DPC相同的结果，对于n = 5个用户，复杂性较低约20 dB。
Dirty Paper Coding (DPC) is considered as the optimal precoding which achieves capacity for the Gaussian Multiple-Input Multiple-Output (MIMO) broadcast channel (BC). However, to find the optimal precoding order, it needs to repeat N! times for N users as there are N! possible precoding orders. This extremely high complexity limits its practical use in modern wireless networks. In this paper, we show the equivalence of DPC and the recently proposed Higher Order Mercer's Theorem (HOGMT) precoding[1][2] in 2-D (spatial) case, which provides an alternate implementation for DPC. Furthermore, we show that the proposed implementation method is linear over the permutation operator when permuting over multi-user channels. Therefore, we present a low complexity algorithm that optimizes the precoding order for DPC with beamforming, eliminating repeated computation of DPC for each precoding order. Simulations show that our method can achieve the same result as conventional DPC with about 20 dB lower complexity for N = 5 users.
