论文标题
Quic-Est：一种传输方案，以最大化多流相关数据流的VOI
QUIC-EST: A Transmission Scheme to Maximize VoI of Multi-Stream Correlated Data Flows
论文作者
论文摘要
新的高级应用程序，例如自动驾驶和触觉通信，需要传输多感官数据，并且需要低延迟和高可靠性。这些应用程序包括。此类服务的现有实施主要依赖于临时计划和发送费率适应机制，该机制直接由应用程序实施并通过UDP运行。在这项工作中，我们提出了一种传输方案，该方案依赖于最近开发的QUIC运输协议的特征，在需要时提供可靠性，并在不损害延迟的情况下提供标准的拥塞控制。此外，我们提出了一个用于传感器数据传输的调度程序，可以随着时间和跨传感器利用相关性。该机制允许应用程序最大化传输数据的信息（VOI）的价值，正如我们通过在两个现实的应用程序方案中所证明的那样。
New advanced applications, such as autonomous driving and haptic communication, require to transmit multi-sensory data and require low latency and high reliability. These applications include. Existing implementations for such services have mostly relied on ad hoc scheduling and send rate adaptation mechanisms, implemented directly by the application and running over UDP. In this work, we propose a transmission scheme that relies on the features of the recently developed QUIC transport protocol, providing reliability where needed, and standardized congestion control, without compromising latency. Furthermore, we propose a scheduler for sensor data transmissions on the transport layer that can exploit the correlations over time and across sensors. This mechanism allows applications to maximize the Value of Information (VoI) of the transmitted data, as we demonstrate through simulations in two realistic application scenarios.
