CIF: Small: Collaborative Research: Synergistic Exploitation of Network Dynamics and Knowledge Heterogeneity in Wireless Networks

无线网络和服务的普及和加速增长在很大程度上取决于无线频谱的可用性和高效利用。不断增长的频谱需求已经将当前的商业无线网络推向了极限，并强调了对无线系统设计的变革性方法的需求。为了满足这一需求，无线网络正在迅速向高密度、用户部署、以积极的频谱重用为特征的异类基础设施发展。这种演进架构可以实现高数据速率，尽管它们必须在存在严重干扰的情况下操作。在传统的系统设计中，干扰被视为一种负外部性，其最终目标是抑制或缓解干扰。该项目开发了通过协同利用反馈、网络动力学和网络知识异质性来拥抱干扰的新方法。这项工作中利用的关键角度是利用干扰作为辅助信息。该项目将描述反馈提供的收益的特征，并分析此类收益的可扩展性和对网络拓扑的依赖。此外，研究人员还调查了网络拓扑中动态变化的影响，并设计了利用这些变化来提高整体网络性能的算法。研究人员还通过考虑网络信道知识在空间和时间域表现出可变性的场景，并开发算法来利用这种可变性，来研究多流多天线无线系统中异质信道知识的最佳利用。

The ubiquity and accelerated growth of wireless networks and services is critically dependent on the availability and efficient use of wireless spectrum. Increasing demand for spectrum is already pushing the current commercial wireless networks to their limits and accentuates the need for transformative approaches for wireless system design. In order to meet this demand, wireless networks are rapidly evolving towards a highly dense, user-deployed, heterogeneous infrastructure characterized by aggressive spectral reuse. Such evolutionary architectures can realize high data rates although they must operate in the presence of severe interference. In traditional system design, interference is viewed as a negative externality with the end goal being its suppression or mitigation. This project develops new approaches for embracing interference through synergistic exploitation of feedback, network dynamics and network knowledge heterogeneity. The key angle leveraged in this work is to exploit interference as side information. The project will characterize the gain provided by feedback, and analyze the scalability and dependence of such gains on network topology. In addition, the researchers investigate the impact of dynamical variations in network topology and devise algorithms that harness such variations to enhance the overall network performance. The researchers also investigate the optimal utilization of heterogeneous channel knowledge for multi-flow multi-antenna wireless systems, by considering scenarios in which network channel knowledge exhibits variability in spatial and temporal domains and developing algorithms to exploit such variability.