The Research Of Active Network Congestion Control Algorithm Based On Operational Data

Abstract

This paper proposes a new congestion control algorithm modified forward active network congestion control algorithm(MFACC), which is based on operational data and designed on the basic of FACC. MFACC applies different processing methods to different kinds of operational data (such as video data, audio data, file data and message data) because of the different characteristics and requests on network resources, adopts the cooperation of the load balanced management algorithm and RED queue management algorithm, and introduces the active detection and the passive indication mechanism. It succeeds in avoiding the algorithm invalidation caused by the indication messages loss, promotes the efficient usage of network resources. A computer simulation is given to analyze the performance of the algorithms. With the analysis of performance between FACC and MFACC, it shows that MFACC not only resolves some problems in FACC, but also improves the OoS of different kinds of operational data, reduces the package loss rate and decreases the processing delay of packets. However, when active node starts just now or the network is idle, many active detection messages will be produced, the network resource will be wasted seriously. At the same time, there is also some limitations in processing video data, because it only aims at MPEG format.