Transmitter- and receiver-based optimization of signaling-assisted modulation classification in MIMO-OFDM systems with adaptive modulation and coding

自适应调制和编码是OFDM系统中一项强大的技术，它可以使发射信号最优地适应频率选择传播信道的当前状态。一个主要的缺点是发射器和接收器必须与所有子载波上使用的调制和编码方案同步。在具有快速变化的传播信道的移动通信系统中，这些传输参数的显式信令-特别是调制方案-会导致有用数据速率的显式降低。或者，可以根据观察到的接收信号自动分类作为子载波调制方案列表的位分配表。本文研究的方法考虑了时分双工系统中自适应传输算法和调制分类的联合优化问题。整个应用项目的目标是开发新的方法来分类多输入多输出（MIMO）系统的数字调制方案，具有自适应调制和编码，以最小的信令开销最大化有效带宽效率方面的系统性能。在第一个资助阶段，新的信令概念已经开发出来，一方面提高了分类的准确性，另一方面与传统的信令概念相比减少了开销。此外，还通过对发射端和接收端策略的联合优化实现了进一步的改进。目前，分类方法已扩展到多天线系统。在第一个项目阶段取得的结果显示了基于信号的调制分类在无线移动通信系统中的巨大潜力，并激发了进一步的研究。在本项目的后三分之一部分，将对最大后验分类算法进行验证，并使用软件定义无线电平台对其参数化进行优化。高分类可靠性的一个关键因素是对互易性的利用，而互易性至少部分存在于所考虑的时分双工系统中。在现实场景中可用的互易量将由实验系统中分类算法的实现来确定。由此，需要遵循必要的信令开销来确保可靠的分类。

Adaptive modulation and coding is a powerful technique in OFDM systems to optimally adapt the transmit signal to the current state of the frequency-selective propagation channel. One major drawback is that transmitter and receiver must be synchronized with respect to the modulation and coding schemes used on all subcarriers. In mobile communication systems with rapidly varying propagation channels, explicit signaling of these transmit parameters - especially of the modulation schemes - causes a significant reduction of the useful data rate. Alternatively, the bit allocation table which is the list of modulation schemes on the subcarriers can be automatically classified based upon the observed receive signal.The approach investigated here considered the joint optimization of the adaptive transmit algorithm and modulation classification in a time-division-duplex system where a minor amount of signaling information is utilized. The objective of the overall applied project is to develop new methods to classify digital modulation schemes for multiple-input multiple-output (MIMO) systems with adaptive modulation and coding which maximize the system performance in terms of the effective bandwidth efficiency with minimal signaling overhead.In the first funded period new signaling concepts have been developed which increase the accuracy of the the classification on the one hand, but reduce the overhead compared to conventional signaling concepts on the other hand. Moreover, further improvements have been achieved by a joint optimization of the transmitter and receiver strategy. Currently, the classification methods are extended to multiple-antenna systems. The results achieved in the first project phase show ernormeous potential of signaling-based modulation classification also in application to wireless mobile communication systems and motivate further investigations.In the last third of the project herewith applied, the maximum-a-posteriori classification algorithms are to be verified and their parametrization is to be optimized using a software-defined-radio platform. One key factor of a high classification reliability turned out to be the utilization of reciprocity which at least partly exists in the considered time-division duplex system. The amount of reciprocity available in realistic scenarios is to be determined by the implementation of the classification algorithms in an experimental system. From that follows the requirements to the necessary signaling overhead to ensure a reliable classification.