Implementation and Evaluation of Video-Quality Adjustment for Heterogeneous Video Multicast Tatsuya Yamada * , Naoki Wakamiya ** , Masayuki Murata ** , and Hideo Miyahara ** * Graduate School of Engineering Science, Osaka University, Japan ** Graduate School of Information Science and Technology, Osaka University, Japan E-mail: t-yamada@ics.es.osaka-u.ac.jp, {wakamiya,murata,miyahara}@ist.osaka-u.ac.jp Abstract By introducing video-quality adaptation mechanisms into intermediate network equipments using active network technologies, we can provide users with video distribution services taking into account client heterogeneity in terms of available bandwidth, performance of client systems, and user’s preferences about video quality. In this paper, we implement the low-pass filter, a quality adjustment technique for real-time multicasting of MPEG-2 video, on an Intel IXP1200 network processor-based network node. We applied the filter to video streams pass- ing through the node and evaluated its practicality and applicability in term of accuracy of video rate adaptation, variation of video quality, and filtering throughput. From the result of evaluation experiments, we demonstrate that the implemented video-quality adjustment mechanism has sufficient rate adaptation capability, and that the low-pass filter is able to accelerate with parallel processing. Keywords: heterogeneous video multicast, video-quality adjustment, network processor, active network 1. Introduction With the proliferation of broadband access to the Internet, video distribution services such as video- streaming or live transmissions are now becoming widely deployed. Since video services involve those users who are heterogeneous in terms of the capacity of access links, the available network bandwidth, the performance of client machines, and the user’s preferences on the perceived video qual- ity, we should introduce mechanisms so that a video stream provided meets user’s environment and preferences. In [1], we proposed mechanisms for video multi- cast services in which diverse client requests are simultaneously satisfied while network resources are efficiently used. Our mechanisms are developed on the basis of active network technologies where intermediated network nodes, called active nodes, adapts the video rate to the desired level. An active network is a network whose behavior can be dynamically and flexibly tailored to network administrator’s, user’s, or even application’s de- mands [2]. Each packet passing through a network equipment, called active node, is processed in ac- cordance with a program that is contained in the packet itself or has been preloaded at the node. By introducing programs to active nodes, they can per- form highly intelligent packet processing from lower-layer functions such as QoS routing to appli- cation-layer functions that manipulate user data in packet payload. In our video multicast mechanisms proposed in [1], to cope with the client-to-client heterogeneity, appropriately chosen active nodes are configured to adapt the rate of an incoming video stream to the desired level by means of video-quality adjustment as illustrated in Fig. 1. In [3], we compared several quality-adjustment mechanisms for real-time MPEG-2 video multicast, namely frame discarding, low-pass, and requantiza- tion filters. We proposed algorithms for those qual- ity-adjustment mechanisms to adapt the video traf- fic to the specified target rate. We conducted sev- eral experiments and concluded that the low-pass filter, which provides rate reduction by progres- sively eliminating high-frequency components of the video signal, is the most effective in terms of suppression of the quality degradation and granu- larity of the rate adaptation. However, we did not consider several implementation-related issues such as per-packet processing, and limitations in the available memory and the processing capability of the network equipment. To evaluate and verify the practicality and appli- cability of the video-quality adjustment within a network, we implemented the proposed mechanism on a network processor-based programmable net- work equipment. In this paper, we show details of Figure 1: Heterogeneous video multicast video low-bandwidth link high quality low-quality video client high-quality video low-quality video low performance high performance client server video quality adjustment