Interactive Distributed Multimedia Systems and Telecommunication Services

Abstract

In this paper, we revisit and challenge the dogma that TCP is an undesirablechoice for streaming multimedia, video in particular. For some time,the common view held that neither TCP nor UDP, the Internet�s maintransport protocols, are adequate for video applications. UDP�s servicemodel doesn�t provide enough support to the application while TCP�sprovides too much. Consequently, numerous research works proposednew transport protocols with alternate service-models as more suitablefor video. For example, such service models might provide higherreliability than UDP but not the full-reliability of TCP. More recently,study of Internet dynamics has shown that TCP�s stature as the predominantprotocol persists. Through some combination of accident and design,TCP�s congestion avoidance mechanism seems essential to the Internet�sscalability and stability. Research on modeling TCP dynamics in orderto effectively define the notion of TCP-friendly congestion avoidanceis very active. Meanwhile, proposals for video-oriented transportprotocols continue to appear, but they now generally include TCP-friendlycongestion avoidance. Our concern is over the marginal benefit ofchanging TCP�s service model, given the presence of congestion avoidance.As a position paper, our contribution will not be in the form offinal answers, but our hope is to convince the reader of the meritin re-examining the question: do applications need a replacementfor TCP in order to do streaming video?