Poster: Link Adaptation in 60 GHz WLANs using PHY Layer Information
Poster: Link Adaptation in 60 GHz WLANs using PHY Layer Information
复制标题
海报：使用 PHY 层信息在 60 GHz WLAN 中进行链路自适应
DOI：
10.1145/3376897.3379168
复制
发表时间：
2020
期刊：
影响因子：
--
通讯作者：
Koutsonikolas, Dimitrios
中科院分区：
文献类型：
--
作者：
Aggarwal, Shivang;Sardesai, Urjit Satish;Sinha, Viral;Koutsonikolas, Dimitrios
速率自适应和波束自适应是60ghz无线局域网中两种主要的链路自适应机制。我们首先使用商品设备展示了，用于确定两种机制中应该首先触发哪一种的简单启发式，即使在看似简单的场景中也可能导致错误的决策。然后，我们使用从软件定义的无线电试验台收集的大型数据集来研究使用几个物理层指标来确定首先触发哪种机制的有效性。最后，我们首次探索了基于机器学习的方法的使用，并表明与COTS设备使用的简单启发式方法相比，它们提供了高精度，并且产生了显着更低的开销和更低的链路恢复延迟。
Rate adaptation and beam adaptation are the two main link adaptation mechanisms in 60 GHz WLANs. We first show, using commodity devices, that simple heuristics, used to determine which of the two mechanisms should be triggered first, can lead to wrong decisions even in seemingly simple scenarios. We then use a large data set collected from a software-defined radio testbed to investigate the effectiveness of using several PHY layer metrics to determine which mechanism to trigger first. We finally explore, for first time, the use of ML-based approaches and show that they offer high accuracy, and incur significantly lower overhead and lower link recovery delay compared to simple heuristics used by COTS devices.