There are many factors governing the performance of TCP-based applications on satellite channels. The end-to-end performance of TCP is known to be degraded by the delay, noise and asymmetry inherent in geosynchronous systems. These results have been based on experiments that evaluate the performance of TCP in single flow tests. Single flow tests are useful for deriving information on the theoretical behavior of TCP and allow for easy diagnosis of problems. However, they do not represent realistic situations and therefore cannot be used to authoritatively comment on performance issues. Our experiments tested TCP's performance in a more dynamic environment with competing traffic flows from hundreds of TCP connections running simultaneously across the satellite channel.
Another aspect investigated was TCP reaction to bit errors on satellite channels. TCP interprets delays and losses as a sign of network congestion. This causes it to cut down its transmission rate leading to poor performance. We allowed the bit error rate on our satellite channel to vary widely and tested the performance of TCP as function of these bit error rates. Our studies show that the average performance of TCP on satellite channels in these realistic conditions is good and establishes the robustness of TCP on satellite channels. Many of our observations contradict commonly held beliefs about the performance of TCP on satellite links.