In this paper, we investigate the joint device activity and data detection in
massive machine-type communications (mMTC) with a one-phase non-coherent
scheme, where data bits are embedded in the pilot sequences and the base
station simultaneously detects active devices and their embedded data bits
without explicit channel estimation. Due to the correlated sparsity pattern
introduced by the non-coherent transmission scheme, the traditional approximate
message passing (AMP) algorithm cannot achieve satisfactory performance.
Therefore, we propose a deep learning (DL) modified AMP network (DL-mAMPnet)
that enhances the detection performance by effectively exploiting the pilot
activity correlation. The DL-mAMPnet is constructed by unfolding the AMP
algorithm into a feedforward neural network, which combines the principled
mathematical model of the AMP algorithm with the powerful learning capability,
thereby benefiting from the advantages of both techniques. Trainable parameters
are introduced in the DL-mAMPnet to approximate the correlated sparsity pattern
and the large-scale fading coefficient. Moreover, a refinement module is
designed to further advance the performance by utilizing the spatial feature
caused by the correlated sparsity pattern. Simulation results demonstrate that
the proposed DL-mAMPnet can significantly outperform traditional algorithms in
terms of the symbol error rate performance.

本文中，我们研究了使用单相不相干方案进行大规模机器型通信（mMTC）中的关节设备活动和数据检测，其中数据位嵌入到导联序列中，基站同时检测活动设备及其嵌入的数据位而无需显式通道估计；针对非相干传输方案引入的相关稀疏性模式，传统的幅度估计传输（AMP）算法无法达到令人满意的性能，因此，我们提出了一种深度学习修改的 AMP 网络（DL-mAMPnet）。DL-mAMPnet 通过有效利用导联活动相关性来增强检测性能。DL-mAMPnet 通过将 AMP 算法展开为前馈神经网络来构建，将 AMP 算法的原理数学模型与强大的学习能力相结合，从而使两种技术的优势得到了发挥。DL-mAMPnet 引入了可训练参数来近似相关稀疏性模式和大规模衰落系数。此外，设计了一种细化模块，通过利用相关稀疏性模式引起的空间特性进一步提高性能。仿真结果表明，所提出的 DL-mAMPnet 在符号误差率性能方面可以显著优于传统算法。

非相干大规模机器通信的模型驱动深度学习

Model-Driven Deep Learning for Non-Coherent Massive Machine-Type  Communications

Massive Machine-Type Communications (MTC) over cellular networks is expected
to be an integral part of wireless "Smart City" applications. The Long Term
Evolution (LTE)/LTE-Advanced (LTE-A) technology is a major candidate for
provisioning of MTC applications. However, due to the diverse characteristics
of payload size, transmission periodicity, power efficiency, and quality of
service (QoS) requirement, MTC poses huge challenges to LTE/LTE-A technologies.
In particular, efficient management of massive random access is one of the most
critical challenges. In case of massive random access attempts, the probability
of preamble collision drastically increases, thus the performance of LTE/LTE-A
random access degrades sharply. In this context, this article reviews the
current state-of-the-art proposals to control massive random access of MTC
devices in LTE/LTE-A networks. The proposals are compared in terms of five
major metrics, namely, access delay, access success rate, power efficiency, QoS
guarantee, and the effect on Human-Type Communications (HTC). To this end, we
propose a novel collision resolution random access model for massive MTC over
LTE/LTE-A. Our proposed model basically resolves the preamble collisions
instead of avoidance, and targets to manage massive and bursty access attempts.
Simulations of our proposed model show huge improvements in random access
success rate compared to the standard slotted-Aloha-based models. The new model
can also coexist with existing LTE/LTE-A Medium Access Control (MAC) protocol,
and ensure high reliability and time-efficient network access.

本篇论文旨在回顾控制 LTE/LTE-A 网络中的大规模机器类型通信 (Massive Machine-Type Communications，MTC) 设备的随机接入的现状和最新提案，提出一种碰撞解决方案的随机接入模型来解决泊松分布访问竞争问题，并相比于标准的基于分时 Aloha 协议的模型，在整体访问成功率和功率效率上取得了显著改进。