Distributed paradigm that aggregates distributed resources to deal with a single task is prevailing over the Internet including parallel transmissions which associate multiple connections to retrieve a single data. In this paper, we propose a parallel transmission scheme for video streaming which receives video data from multiple senders simultaneously including several features to resolve the challenges involved when multiple connections are implemented with video streaming. In order to satisfy the progressive-transmission requirement of video streaming as well as to secure the real-time inter-operability among multiple connections, we introduce a packet-by-packet flow aggregation framework and a centralized scheduler based rate allocation. We also introduce a real-time broken link detection scheme using a virtual measure called time-queue and implement retransmissions in order to distribute faults of links among the participating parallel connections. Through the simulation we verify that the proposed scheme greatly improves robustness and stability of transmission channels which is one of key requirements of video streaming. The rate shared by broken connections are successfully compensated by other healthy connections in real-time and a seamless playback is guaranteed even over a severe degradation of individual connections.
In addition, we also propose a receiver buffer control mechanism that uses a PI-type feedback control which takes advantage of the robust channel characteristics of the parallel streaming. In a single connection based streaming, people tried to make the output rate of media to be adapted to the available rate of Internet connections because there is no way to reserve the affordable rate for streaming. However, by using multiple connections, the transmission channel can respond flexibly to the rate changes brought by the feedback control and it enables even VBR-encoded media to be streamed without any a priori information about the media. A mathematical fluid model is set up to analysis and find the stable region of the control system and we determine the control gains through the analysis. The simulation results prove that this control successfully maintains the level of the receiver buffer regardless the streaming media was encoded with CBR or VBR.
Finally, in order to take the wireless channels around the Internet into consideration, we pro-pose a remedy to get around the problem of discrimination between wireless losses and congestion losses. In a wireless network packet losses can be caused not only by network congestion but also by unreliable error-prone wireless links. Therefore, flow control schemes which use packet loss as a congestion measure cannot be directly applicable to a wireless network because there is no way to distinguish congestion losses from wireless losses. In order to get around this problem we use ECN (Explicit Congestion Notification) marking in conjunction with RED (Random Early Detection) queue management scheme intelligently. By this method, not only the degree of network congestion is notified to multimedia sources explicitly in the form of ECN-marked packet probability but also wireless losses are hidden from multimedia sources. We calculate TCP-friendly rate based on ECN-marked packet probability instead of packet loss probability, thereby effectively eliminating the effect of wireless losses in flow control and thus preventing throughput degradation of multimedia flows travelling through wireless links. Through the simulations we show that the proposed scheme indeed improves the quality of delivered video significantly while maintaining TCP-friendliness in a wireless environment.
