Advanced Hybrid Analog-Digital Massive MIMO Techniques for Millimeter Wave Wireless Systems (AdAMMM)

与6 GHz以下稀缺的无线电资源相比，毫米波(MmWave)频率下的大量可用频谱具有巨大的潜力来满足下一代无线通信系统的通用和关键需求。然而，毫米波信号遭受更大数量级的路径损耗，并且由于笔状波束宽度而具有更高的阻塞概率。这些缺点可以通过在发射机和/或接收机使用大量天线来补偿，即通过使用大规模MIMO来补偿。一种实用化的大规模MIMO结构是模拟-数字混合MIMO结构，它是一种成本和能源效率高的结构。它可以从使用模拟和数字处理中受益。模拟部分通常使用移相网络来实现。这种新的硬件结构对MIMO处理算法施加了新的限制，并带来了新的信号处理挑战。因此，先进的模拟-数字混合大规模MIMO算法将成为毫米波无线通信的关键使能。本项目的目的是研究和评估模拟-数字混合大规模MIMO技术在未来宽带毫米波无线通信中的潜力。将对混合MIMO辅助毫米波无线通信的各个方面进行研究和评估。这将通过开发高效的混合MIMO预编译码算法以及信道估计算法来实现，并通过推导相应的信息论结果来实现。我们将研究点对点场景、多用户MIMO场景和多用户多小区场景，同时考虑时变或时不变的频率选择性信道模型。为了提高覆盖范围，将开发混合MIMO双向中继技术。混合MIMO架构的性能也将与竞争对手基于1位模数转换器的大规模MIMO架构进行比较。

Compared to the scarce radio resources under 6 GHz, large available portions of spectrum at millimeter wave (mmWave) frequencies have a great potential to satisfy versatile and critical demands of next generation wireless communication systems. However, mmWave signals suffer an orders-of-magnitude larger path loss and have a higher probability of blockage due to the pencil-like beamwidth. These drawbacks can be compensated by using a large number of antennas at the transmitter and/or at the receiver, i.e., by using massive MIMO. One of the practical massive MIMO architectures, which is cost and energy efficient, is a hybrid analog-digital MIMO architecture. It can benefit from the use of both analog and digital processing. The analog parts are often realized using phase shift networks. This new hardware architecture imposes new constraints on MIMO processing algorithms and introduces novel signal processing challenges. Therefore, advanced hybrid analog-digital massive MIMO algorithms, which tackle these new challenges, will be a key enabler of mmWave wireless communications.The aim of this project is to study and to evaluate the potential of hybrid analog-digital massive MIMO techniques for future wideband mmWave wireless communications. Various aspects of hybrid MIMO aided mmWave wireless communications will be investigated and evaluated. This will be achieved by developing efficient hybrid MIMO precoding and decoding algorithms as well as channel estimation algorithms, and by deriving the corresponding information theoretic results. We will study the point-to-point scenario, multi-user MIMO scenarios and multi-user multi-cell scenarios taking into account time-varying or time-invariant frequency-selective channel models. To improve the coverage, hybrid MIMO two-way relaying techniques will be developed. The performance of hybrid MIMO architectures will also be compared to the competing massive MIMO architecture based on 1-bit analog-to-digital converters.