Research And Simulation Of Queue Management Mechanism In Context-Aware Autonomic Network

While the quickly increasement of network services and rapidly expansion of network scale, the drawbacks of network architecture are also becoming obvious. So there is an urgent need to improve it. More and more researchers have tried to carry out research on new network architecture. As one research direction of new network architecture, autonomic network has been receiving more and more attention.Emerging network applications'requirements of quality of service are increasingly high, so it is increasingly important to use congestion control in network. As a main mechanism supporting congestion control at routers, active queue management obtained research widely. However, the majority of active queue management mechanism has been only designed for TCP, and does not apply to the use of UDP. Packet loss caused by network congestion has severe affect on the multimedia applications using UDP protocol. UDP doesn't have congestion control mechanism at the network layer, the transmitter can not be informed of the occurrence of congestion and still send data with the original rate, resulting in a large number of data loss. Therefore, the current queue management mechanism can not effectively support the quality of multimedia applications.At present, few researchers study the queue management mechanism to support multimedia streaming's quality of service. The bandwidth required for multimedia streaming is wide, it is a challenge for the wireless environment to support multimedia applications, where the packet is easy to loss and the resource is constrained. In order to provide quality of service for multimedia applications in wireless environment, we study and propose a queue management mechanism based on context aware, named AAQM (Autonomic Active Queue Management). It is capable of adaptively adjust the packet loss operation by the perceived network and service context information, to provide better perceived video quality.