A Video Transmission Framework Based on Packet Criticality Model

During video streaming over packet-erasure network, packets may lose due to network congestion. Since video packets have different significance and validity for decoding, the streaming system is expected to transmit the most critical packets when congestion occurs. In this paper, a packet criticality model is proposed and then a video transmission framework based on this model is presented. The criticality model describes how critical a video packet should be sent and provides information to transmitter to decide which packet to be transmitted firstly. Experimental results show that the proposed criticality model can help video streaming system with improved performance and raised quality of reconstructed images.