In this article, a novel multiple-input multiple-out (MIMO) transmission scheme, called generalized polarized enhanced spatial modulation (GPESM), is proposed for dual-polarized land mobile satellite (LMS) communications. We first introduce the enhanced spatial modulation (ESM) technique for dual-polarized LMS communications, in which polarization dimension, spatial dimension and multiple signal constellations are used to transmit information and obtain substantial performance gain. Meanwhile, the theoretical upper bound for the average bit error probability (ABEP) of the proposed GPESM scheme is derived. In order to further improve the reliability of the system, we also propose two novel power allocation (PA) algorithms for GPESM system, which are the optimization-driven approximated max-min distance (AMMD)-based PA algorithm and the data-driven deep neural network (DNN)-based PA algorithm. To achieve an enhanced spatial diversity gain, we consider to apply a reconfigurable intelligent surface (RIS) to the GPESM system as a relay to assist in transmitting information. In this way, the user can receive the information transmitted by the satellite on one hand, and the information sent by the satellite via the RIS relay on the other hand. We also extend the above-mentioned two PA algorithms to the RIS-assisted GPESM systems. Our simulation results show that the RIS-assisted GPESM systems are capable of obtaining high bit error rate (BER) performance gain (up to 10 dB) compared to the standard GPESM system and two PA algorithms can further improve the performance to the systems.

中文翻译：

---

陆地移动卫星通信的多维偏振调制


在本文中，提出了一种用于双极化陆地移动卫星（LMS）通信的新型多输入多输出（MIMO）传输方案，称为广义极化增强空间调制（GPESM）。我们首先介绍用于双偏振LMS通信的增强型空间调制（ESM）技术，其中偏振维度、空间维度和多个信号星座用于传输信息并获得显着的性能增益。同时，推导了所提出的GPESM方案的平均误码概率（ABEP）的理论上限。为了进一步提高系统的可靠性，我们还为GPESM系统提出了两种新颖的功率分配（PA）算法，即基于优化驱动的近似最大最小距离（AMMD）的PA算法和数据驱动的深度PA算法。基于神经网络 (DNN) 的 PA 算法。为了实现增强的空间分集增益，我们考虑将可重构智能表面（RIS）应用于GPESM系统作为中继来协助传输信息。这样，用户一方面可以接收卫星发送的信息，另一方面可以通过RIS中继接收卫星发送的信息。我们还将上述两种 PA 算法扩展到 RIS 辅助的 GPESM 系统。我们的仿真结果表明，与标准 GPESM 系统相比，RIS 辅助的 GPESM 系统能够获得高误码率 (BER) 性能增益（高达 10 dB），并且两种 PA 算法可以进一步提高系统性能。