A fuzzy approach to MPEG video transmission in ATM networks

This paper is concerned with transmission of moving picture expert group (MPEG) variable bit rate (VBR) data sources over asynchronous transfer mode (ATM) networks, using a self-organising fuzzy (SOF) controller and a rule-based fuzzy (RBF) controller. The variations in bit rate for MPEG VBR data sources are predominantly large and unpredictable, which could cause data loss and subsequently image quality degradation in ATM networks. Conventional rate policing such as generic cell rate algorithm and traffic-shaping buffer are largely inadequate to sufficiently regulate transmission of VBR data sources over ATM. In this research, a new SOF controller is developed to eliminate excessive delay or loss of the VBR data sources at the user-network-interface. The SOF controller adjusts the traffic-shaping buffer output rate frame by frame to enable the VBR encoded video to conform to the leaky-bucket's contract prior entering ATM. A RBF controller is used to police the data rate entering the traffic shaper, in order to prevent either overspill or starvation of the buffer on a group of picture (GOP) by GOP basis. The simulation results show that the use of the SOF controller and the RBF controller reduce delay or data loss at the user-network-interface compared with a conventional rate policing mechanism in ATM.