Streaming media congestion control using bandwidth estimation

The fundamental challenge in streaming media over the Internet is to transfer the highest possible quality, adhere to the media play out time constraint, and efficiently and fairly share the available bandwidth with TCP, UDP, and other traffic types. This work introduces the Streaming Media Congestion Control protocol (SMCC), a new adaptive media streaming congestion management protocol in which the connection's packet transmission rate is adjusted according to the dynamic bandwidth share of the connection. In SMCC, the bandwidth share of a connection is estimated using algorithms similar to those introduced in TCP Westwood. SMCC avoids the Slow Start phase in TCP. As a result, SMCC does not exhibit the pronounced rate oscillations characteristic of traditional TCP, thereby providing congestion control that is more suitable for streaming media applications. Furthermore, SMCC is fair, sharing the bandwidth equitably among a set of SMCC connections. An important advantage is robustness when packet losses are due to random errors, which is typical of wireless links and is becoming an increasing concern due to the emergence of wireless Internet access. In the presence of random errors, SMCC is also friendly to TCP New Reno. We provide simulation results using the ns2 simulator for our protocol running together with TCP New Reno.