作者: Ayon Chakraborty , Shruti Sanadhya , Samir R. Das , Dongho Kim , Kyu-Han Kim
关键词:
摘要: Enterprise wireless networks face significant challenges to deliver Quality-of-Experience (QoE) with the variety of mobile applications. One fundamental is that traditional definition network capacity (often defined as throughput capacity) not sufficient reflect applications' requirements in networks. In this paper, we propose rethink better incorporate QoE. Specifically, first a novel concept an Experiential Capacity Region (ExCR) for ExCR set simultaneous application flows whose QoE can be satisfied by network. Next, present infrastructure based ExBox system measures per-application metrics and determines serve flows. its core, employs light-weight machine learning techniques are tailored dynamic environments. Through both large-scale simulations extensive real-life experiments on WiFi LTE networks, show delivers admission control decision precision ≈ 0.8 - 0.9, even when clients experience diverse channel quality. Moreover, quickly adapts changing environments without much overhead.