Scalability improvement of the real-time control protocol (RTCP) leading to management facilities in the Internet

Abstract

We address some of the problems concerning RTCP scalabilty in large multicast groups. The problems are summarised as the increased storage state maintained by every member about every other member in the group, the feedback delay, and the bandwidth usage problem especially for receivers that obtain incoming RTCP reports through low bandwidth links. We designed a scheme to tackle these problems and to improve the RTCP scalability. Our scheme is based on a hierarchical structure in which members are grouped dynamically, into local regions and receiver reports (RRs) are sent with limited scope. As a result, the load is distributed among members in the group. Each region has an aggregator (AG) that receives the RRs feedback, performs some aggregation statistics on this control data, then sends these statistics to a manager. The manager performs additional analysis to evaluate the network performance and to estimate regions which are suffering from significant congestion. Finally, we present the benefits of using our scheme.