Abstract:
The Internet is experiencing massive growth in multimedia applications like streaming audio and video. These applications normally have no transmission rate control mechanism as a result they congest a bottleneck link within the Internet. Therefore, it is very important that all applications implement some form of congestion control. To address these maladies we proposed a novel, reliable and preemptive Dynamic Congestion Control Mechanism (DCCM) based on RTP for streaming media.
The basic approach adopted is decoupling of persistence congestion from transient congestion. Our mechanism is tailored to use packet inter-arrival jitter for detecting transient congestion and estimated packet loss rate probability for detecting and persistence congestion. The jitter sensor approach used in Dynamic Congestion Control Mechanism (DCCM) anticipates congestion and dynamically regulate transmission rate according to network condition before packet losses occur. It is well suited for real-time streams and other semi reliable rate based applications. The primary goal is to detect congestion before it leads to packet losses while trying to maintain fairness with other competing flows. Maintaining a steady behavior , smooth transmission rate and efficient utilization of available bandwidth in a scalable manner is also ensured in this model ,a major QoS factor(s) for multimedia communication.
Dynamic Congestion Control Mechanism (DCCM) have been evaluated and tested through simulation and concludes that bandwidth is almost evenly shared between Dynamic Congestion Control Mechanism (DCCM) traffic and TCP flow ,while Dynamic Congestion Control Mechanism (DCCM) tries to maintain smoothness by effective utilization of bandwidth. Unlike most other congestion control mechanisms which rely solely on packet drop for detecting congestion , Dynamic Congestion Control Mechanism (DCCM) is able to prevent congestion by anticipating jitter in network. Dynamic Congestion Control Mechanism (DCCM) regulate its transmission rate dynamically by monitoring current congestion level of network, thus intelligently prevents congestion while maintaining steady rate behavior.
