arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2002.10965eess.SP
收藏

IRS辅助的大规模被动恒包络预编码MIMO系统

IRS-aided Large-Scale MIMO Systems with Passive Constant Envelope Precoding

Yasaman Omid, Seyyed MohammadMahdi Shahabi, Cunhua Pan, Yansha Deng, Arumugam Nallanathan

摘要:

本文研究了一种智能反射面辅助的大规模MIMO系统,在该系统中,每个基站(BS)均采用恒包络预编码(CEP)。在大规模天线阵列中实现CEP既节省成本又节省能量。我们的目标是通过我们提出的三种算法来优化BS和IRS的离散相移,以最小化系统中多用户干扰(MUI)的总功率。为了简单起见,考虑简单的单小区场景,其中BS和IRS相移的优化通过低复杂度的基于网格的算法来解决。然后,该算法被扩展到多小区场景,其中每个BS中的预编码操作被单独执行。借助随机优化方法,提出了一种低开销的基于网格的解决方案,该方案比第一种方案具有更好的性能。最后,我们通过半定松弛(SDR)计划解决优化问题,作为所提出的算法的性能基准。同时,对提出的算法进行了干扰分析和复杂度分析.数值结果表明,虽然网格算法的性能略低于连续相位SDR算法,但前者的计算复杂度和实现成本远低于后者,具有较好的实际应用价值.

英文摘要:

In this paper, an intelligent reflecting surface (IRS)-aided large-scale MIMO system is investigated in which constant envelope precoding (CEP) is utilized at each base station (BS). It is both cost-effective and energy-efficient to implement CEP in a large-scale antenna array. We aim to optimize the discrete phase shifts at both the BS and the IRS to minimize the sum power of multi-user interference (MUI) in the system via our proposed three algorithms. For the sake of simplicity, a simple single-cell scenario is considered, where the optimization of the BS and IRS phase shifts is solved by a low-complexity trellis-based algorithm. Then, this algorithm is extended to a multi-cell scenario, where the precoding operation in each BS is performed individually. With the aid of stochastic optimization method, a low-overhead trellis-based solution is proposed which has better performance than the first one. Finally, we solve the optimization problem via the semi-definite relaxation (SDR) scheme, to serve as a performance benchmark for the proposed algorithms. Meanwhile, interference and complexity analysis is provided for the proposed algorithms. Numerical results demonstrate that while the performance of the trellis-based algorithms is negligibly lower than that of the continuous-phase SDR-based solution, the computational complexity and the implementation cost of the former is much lower than the latter, which is appealing for practical applications.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑