Performance Analysis and Optimization of Future Low Latency and Reliable Cooperative Communication

随着低延迟和高可靠性成为未来无线网络设计的两大重点，各种技术被提出、研究和改进以满足这些需求。其中，中继作为一种利用空间分集来缓解无线衰落的有效方法而广为人知。然而，几乎所有现有的关于中继优势的研究都是在理想的假设下进行的，即在香农的信道容量下进行任意可靠的通信，即假设所谓的无限块长度机制。由于低延迟需要较短的分组长度，从而引入译码错误并降低可靠性，因此研究有限分组长度下的中继行为对于设计未来的低延迟、可靠的网络是非常必要的。在该方案中，我们旨在通过利用有限块长度机制的新的性能模型和优化技术来为低延迟和可靠的中继系统的性能分析和系统设计做出贡献。更重要的是，我们考虑了更现实的系统模型，例如有限块长、不完善的CSI和衰落信道模型，这使得我们可以从我们的结果中推导出更合理、更可靠的指导方针，在实践中进行系统设计。首先，我们将对单中继网络在上述现实系统模型下的性能进行详细分析。然后，我们将单中继网络的结果推广到多中继网络。最后，通过利用所获得的有限块长度中继的分析见解，我们打算通过最小化差错概率和最大化服务质量约束的容量来设计有效的服务质量约束的可靠的中继网络。

As low-latency and high reliability become two major concerns in the design of future wireless networks, various technologies are proposed, studied and improved to satisfy these requirements. Among them, relaying is well known as an efficient way to mitigate wireless fading by exploiting spatial diversity. However, almost all the existing studies of the advantages of relaying are under the ideal assumption of communicating arbitrarily reliably at Shannon's channel capacity, i.e. assuming the so-called infinite blocklength regime. As low-latency requires short blocklengths which introduce decoding errors and reduce the reliability, it is essential to study the behavior of relaying in the finite blocklength regime to contribute to the design of future low-latency and reliable networks. In this proposal, we aim at contributing to the performance analysis and system design of low latency and reliable relaying systems by utilizing a novel performance model of the finite blocklength regime and optimization techniques. More importantly, we consider more realistic system models, e.g. finite blocklength, imperfect CSI and a fading channel model, which allows us to deduce from our results more reasonable and reliable guidelines for the system design in practice. First, we will provide a detailed analysis of the performance of single-relay networks under the above described realistic system models. Then, we will extend the obtained results for single-relay networks to multi-relay networks. Finally, by exploiting the obtained analytical insights into relaying with finite blocklengths, we intend to design efficient QoS-constrained and reliable relaying networks by minimizing the error probability and maximizing the QoS-constrained capacity.