The use of a server-based loss recovery technique for reliable multicast can significantly improve the system performance in terms of loss recovery latency and bandwidth consumption. Appropriate congestion control mechanisms are necessary to provide fairness and to maintain a high netWork throughput and link utilization. In reliable multicast protocol design the loss recovery associated with data delivery and congestion control are not independent issues and should be addressed simultaneously. This suggests that a server-based loss recovery technique allied withcongestion control could have potential in addressing the challenges of scalabililty and heterogeneity for reliable multicast in a best-effort netWork. The authors propose a new framework which jointly performs local delivery and congestion control (LDCC). In this framework, delivery and control servers (DCSs) collocated with routers perform LDCC functions. Each DCS and its serving receivers form a local DCS region according to a tree topology. Withproper acknowledgment processing and buffer management, packet loss can be efficiently recovered locally. Also, the overall throughput degradation caused by the interference of neighbouring regions can be mininised by local congestion control. NS-2 simulations are used to demonstrate that the framework can achieve a significantly lower loss recovery latency without sacrificing the network throughput, compared to existing approaches such as AER/NCA. It is also shown using fairness tests that the proposed framework is TCP-compatible.
展开▼
