Efficient Frame Schedule Scheme for Real-time Video Transmission Across the Internet Using TCP Yonghua Xiong School of Information Science and Engineering, Central South University, Changsha, China Email: yhxiong@mail.csu.edu.cn Min Wu 1 , Weijia Jia 2 1 School of Information Science and Engineering, Central South University, Changsha, China 2 Department of Computer Science, City University of Hong Kong, Hong Kong, China Email: min@mail.csu.edu.cn, itjia@cityu.edu.hk Abstract—The great end-to-end delays are the major factor to influence the visual quality of real-time video across the Internet using TCP as transport layer protocol. In this paper, we present a video frame schedule scheme for rate adaptive real-time video transmission over TCP. The scheme schedules video frames between the application layer sender-buffer, the TCP sender-buffer and TCP receiver-buffer and can regulate automaticlly the video frame rate and play out buffer delays according to the network congestion level. The sheme requires only an extra buffer of application layer and can significantly cut down the end-to-end delays of real-time video without any modification to the network infrastructure or TCP protocol stack. The performance of the proposed solution is evaluated through extensive simulations using the NS-2 simulator. Index Terms—TCP, rate adaptive scheme, real-time video, frame schedule, transmission I. INTRODUCTION TCP integrates the flow control, congestion control, and retransmission mechanisms[1], therefore it is well recognized that the AIMD behavior and retransmission timeouts of current TCP would introduce extra delays and rate fluctuation for using TCP to transmitting real-time video flows[2], which could cut down the quality of service of video. Hence, there are many issues that modify TCP to make it more available to transmit video streaming [3][4]. Despite the numerous new algorithms, there is no consensus on the new ultimate transport protocol that would replace the traditional TCP[5]. In current Internet, more than 90% applications use still common TCP version protocol[6]. In addition, due to the obvious advantages of using TCP such as rapid response to network congestion, TCP friendliness, reliable transmission, well-developed, extensively-used, acceptable by firewall and so on, real- time multimedia streaming applications are increasingly using TCP as their underlying transport protocol[7]. Ref [8] proposes a playback adaptation algorithm for video streaming with TCP in wireless networks. Ref[9] propose a scalable wireless video streaming system adopting TCP transmission mode. Transport video streaming uses multiple TCP connections in [10] and [11]. In [12] and [13], rate adaptation is based on the periodic feedback from the client and only applies to layered video. Ref [14] presents a receiver-based bandwidth sharing system for allocating the capacity of last-hop access links according to user preferences. In [15], frame rate is adjusted at the receiver to maximize the visual quality based on the overall loss. In [16], packet dispersion is measured at the receiver to provide a graded way of estimating congestion, and sending rate is controled by the fuzzy controller. In [17], a framework is presented based on cross layer feedback for smoother rate control of streaming. In [7], a bandwidth prediction methodology is established based on TCP time series creation for real-time streaming applications. All of the above schemes, however, have some limitations. The multiple TCP connection method make implementation and maintainance complex. The prioritization of video frames requires extra work at the application level and may not be suitable for live streaming. The schemes based on layered videos restrict the formats of the videos that can be streamed using TCP. The receiver-based scheme requires much timely feedback from receiver and is difficult to be deployed. The bandwidth prediction methodology comparatively depends on the accuracy of prediction model. 216 JOURNAL OF NETWORKS, VOL. 4, NO. 3, MAY 2009 © 2009 ACADEMY PUBLISHER