The performance of TCP/IP for networks with high bandwidth-delay products and random loss

TLDR

This paper examines the performance of TCP/IP over wide‑area networks where data traffic may coexist with real‑time traffic such as voice and video. The study aims to develop a basic understanding of TCP/IP behavior under high bandwidth‑delay product and random loss conditions, and to suggest transport and network layer modifications for improved performance. The authors analyze and simulate TCP/IP behavior in high bandwidth‑delay product networks with random loss. They find that random loss severely degrades throughput when loss probability times the square of the bandwidth‑delay product exceeds one, that TCP is unfair to high‑delay connections sharing a bottleneck, that FIFO queueing may be inadequate, and that Reno is less robust than Tahoe under closely spaced losses.

Abstract

This paper examines the performance of TCP/IP, the Internet data transport protocol, over wide-area networks (WANs) in which data traffic could coexist with real-time traffic such as voice and video. Specifically, we attempt to develop a basic understanding, using analysis and simulation, of the properties of TCP/IP in a regime where: (1) the bandwidth-delay product of the network is high compared to the buffering in the network and (2) packets may incur random loss (e.g., due to transient congestion caused by fluctuations in real-time traffic, or wireless links in the path of the connection). The following key results are obtained. First, random loss leads to significant throughput deterioration when the product of the loss probability and the square of the bandwidth-delay product is larger than one. Second, for multiple connections sharing a bottleneck link, TCP is grossly unfair toward connections with higher round-trip delays. This means that a simple first in first out (FIFO) queueing discipline might not suffice for data traffic in WANs. Finally, while the Reno version of TCP produces less bursty traffic than the original Tahoe version, it is less robust than the latter when successive losses are closely spaced. We conclude by indicating modifications that may be required both at the transport and network layers to provide good end-to-end performance over high-speed WANs.