The large scale multiuser multiple input multiple output (MU-MIMO) is one of the promising communication technology for 5G wireless networks as it offers reliability, high spectral efficiency and high throughput. The lattice reduction (LR) precoding based user level local likelihood ascent search (ULAS) detection scheme is proposed in this paper for efficient signal detection in large scale MU-MIMO system. The initial solution of ULAS algorithm is obtained from the LR precoding assisted zero forcing detector. The LR precoding transforms the non-orthogonal channel matrix into nearly orthogonal channel, which helps to mitigate inter antenna interference (IAI) exists at each user. The remaining multiuser interference (MUI) imposed to each user from undesired users is cancelled by the proposed ULAS multiuser detection scheme. Thus, the proposed LR precoding assisted ULAS mitigates both IAI and MUI unlike the classical detector, those try to moderate either IAI or MUI. By contrast, the proposed ULAS detector provides performance close to optimal maximum likelihood detector with just a fraction of its complexity.

大规模多用户多输入多输出 (MU-MIMO) 是 5G 无线网络的有前途的通信技术之一，因为它具有可靠性、高频谱效率和高吞吐量。本文提出了基于格约化 (LR) 预编码的用户级局部似然上升搜索 (ULAS) 检测方案，用于大规模 MU-MIMO 系统中的有效信号检测。ULAS算法的初始解是从LR预编码辅助迫零检测器中获得的。LR 预编码将非正交信道矩阵转换为近似正交的信道，这有助于减轻每个用户存在的天线间干扰 (IAI)。所提出的 ULAS 多用户检测方案消除了来自不想要的用户对每个用户施加的剩余多用户干扰 (MUI)。因此，与经典检测器不同，所提出的 LR 预编码辅助 ULAS 减轻了 IAI 和 MUI，后者试图缓和 IAI 或 MUI。相比之下，所提出的 ULAS 检测器提供的性能接近于最佳最大似然检测器，而其复杂性只是其一小部分。