In this work, we consider the use of model-driven deep learning techniques for massive multiple-input multiple-output (MIMO) detection. Compared with conventional MIMO systems, massive MIMO promises improved spectral efficiency, coverage and range. Unfortunately, these benefits are at the expense of significantly increased computational complexity. To reduce the complexity of signal detection and guarantee the performance, we present a learned conjugate gradient descent network (LcgNet), which is constructed by unfolding the iterative conjugate gradient descent (CG) detector. In the proposed network, instead of calculating the exact values of the scalar step-sizes, we explicitly learn their universal values. Also, we can enhance the proposed network by augmenting the dimensions of these step-sizes. Furthermore, in order to reduce the memory costs, a novel quantized LcgNet is proposed, where a low-resolution nonuniform quantizer is used to quantize the learned parameters. The quantizer is based on a specially designed soft staircase function with learnable parameters to adjust its shape. Meanwhile, due to fact that the number of learnable parameters is limited, the proposed networks are relatively easy to train. Numerical results demonstrate that the proposed network can achieve promising performance with much lower complexity.

中文翻译：

---

用于大规模 MIMO 检测的学习共轭梯度下降网络

在这项工作中，我们考虑使用模型驱动的深度学习技术进行大规模多输入多输出 (MIMO) 检测。与传统 MIMO 系统相比，大规模 MIMO 有望提高频谱效率、覆盖范围和范围。不幸的是，这些好处是以显着增加计算复杂性为代价的。为了降低信号检测的复杂性并保证性能，我们提出了一个学习共轭梯度下降网络（LcgNet），它是通过展开迭代共轭梯度下降（CG）检测器构建的。在提议的网络中，我们不是计算标量步长的确切值，而是明确地学习它们的通用值。此外，我们可以通过增加这些步长的维度来增强提议的网络。此外，为了降低内存成本，提出了一种新颖的量化 LcgNet，其中使用低分辨率非均匀量化器来量化学习到的参数。量化器基于专门设计的软阶梯函数，具有可学习的参数来调整其形状。同时，由于可学习参数的数量有限，所提出的网络相对容易训练。数值结果表明，所提出的网络可以以低得多的复杂性实现有希望的性能。