Novel congestion control algorithms, like Google’s TCP BBR, have been designed using wired-bottleneck models. Therefore, they do not achieve efficient performance over Wi-Fi paths due to frame-aggregation mechanisms adopted by standard IEEE 802.11n/ac/ax technologies that break the standard model. In this work, we analyze the performance characteristics of a Wi-Fi bottleneck, investigating the throughput and the latency obtained with the frame aggregation logic. In particular, we formulate the expression for data throughput and RTT as a function of the aggregation size. Then, through experimental results, we validate the model on a real testbed with two different types of Wi-Fi bottlenecks.

中文翻译：

---

考虑吞吐量和延迟的 Wi-Fi 帧聚合性能模型

新的拥塞控制算法，如 Google 的 TCP BBR，是使用有线瓶颈模型设计的。因此，由于打破标准模型的标准 IEEE 802.11n/ac/ax 技术采用的帧聚合机制，它们无法在 Wi-Fi 路径上实现高效性能。在这项工作中，我们分析了 Wi-Fi 瓶颈的性能特征，研究了通过帧聚合逻辑获得的吞吐量和延迟。特别是，我们将数据吞吐量和 RTT 的表达式制定为聚合大小的函数。然后，通过实验结果，我们在具有两种不同类型 Wi-Fi 瓶颈的真实测试平台上验证模型。