Optimized Non-Regenerative MIMO Relaying (ONEMORE)

MIMO中继网络模型在未来移动的通信的研究和应用中得到了广泛的应用。这为研究一般MIMO中继网络的基本极限提供了动力。此外，自由度（DoF）是网络的一个重要指标，在高信噪比（SNR）的情况下，它与和速率呈线性关系。在本课题中，我们计划通过理论分析和有效的算法来优化有限信噪比工作点下非再生MIMO中继网络的可达和速率和可达自由度，首先建立了两类模型：带功率约束的和速率最大化模型和带速率约束的和发射功率最小化模型。系统优化的主要困难是和率表达式复杂且高度非线性。在本项目中，将采用几种方法近似计算和率，并对它们的结果进行比较。应用非线性优化技术解决由此产生的非凸非线性矩阵优化问题。将分析算法的计算复杂度和信令开销。基于模型和相应的算法，我们将得到系统的可达速率域和功率域。为了实现高的和率，我们计划设计两种类型的数据流选择算法：一系列的非线性优化子问题的优化技术方法和启发式贪婪方法与流选择标准。本课题的研究工作不仅可以丰富理论研究成果，而且可以为非再生MIMO中继网络的应用提供指导。

MIMO relay network models have been widely used in researches and applications for future mobile communications. This gives an impetus to the investigations of the fundamental limits of the general MIMO relay networks. Also, degree of freedom (DoF) is an important index of a network, which scales linearly with the sum rate in high signal-to-noise-ratio (SNR) scenarios. In our project, we plan to optimize the achievable sum rate and the achievable DoFs of non-regenerative MIMO relay networks for finite SNR operating points, through both theoretical analysis and efficient algorithms.First, we build up two types of models: sum rate maximization models with power constraints and sum transmit power minimization models with rate constraints. The main difficulty to optimize the system is that the sum rate expression is complicated and highly nonlinear. In the project, the sum rate will be approximated by several approaches, and their results compared. Nonlinear optimization techniques are applied to solve the resulting nonconvex nonlinear matrix optimization problems. The computational complexity and their signalling overhead of the algorithms will be analyzed. Based on the models and the corresponding algorithms, we will obtain the system achievable rate region and the power region. To achieve high sum rate, we plan to design two types of data stream selection algorithms: the optimization technique approach with a series of nonlinear optimization subproblems and the heuristic greedy approach with stream selection criteria. The works of our project will not only enhance the theoretical results, but also provide guidelines to applications of non-regenerative MIMO relay networks.