This paper proposes an application-layer control on the number of concurrent senders in an incast application. The ideal number of concurrent senders is computed using the SRU, RTT and bottleneck buffer size. The paper assumes these parameters are available to the incast application.

## Bibliographic data

@inproceedings{
title = "An Effective Approach to Preventing TCP Incast Throughput Collapse for Data Center Networks",
author = "Hongyun Zheng and Chunming Qiao",
booktitle = "Proc GLOBECOM",
month = "Dec",
year = "2011",
}