IEEE JOURNAL ON SELECTED AREAS IN COMMUNICATIONS, VOL. 15, NO. 1, JANUARY 1997 119 MPEG 2 Video Services for Wireless ATM Networks Jian Zhang, Student Member, IEEE, Michael R. Frater, Member, IEEE, John F. Arnold, Senior Member, IEEE, and Terence M. Percival Abstract—Audio-visual and other multimedia services are seen as an important source of traffic for future telecommunications networks, including wireless networks. In this paper, we examine the impact of the properties of a 50 Mb/s asynchronous transfer mode (ATM)-based wireless local-area network (WLAN) on mov- ing picture experts group phase 2 (MPEG 2) compressed video traffic, with emphasis on the network’s error characteristics. The paper includes a description of the WLAN system used and its loss characteristics, a brief discussion of relevant aspects of the MPEG 2 standards and the associated error resilience techniques for minimizing the effect of transmission errors, and a description of the method by which the video data is organized for transmission on the network. We show results on the effect of cell loss due to transmission errors on the quality of the decoded video at the receiver, and demonstrate how error resilience techniques in both the systems and video layers of MPEG 2 can be used to improve the quality of service. Situations where up to 1% of the data is lost due to network transmission errors are examined. Most important among the findings are that error resilience experiments that do not take into account the effect of the MPEG 2 systems layer will tend to significantly overestimate the quality of received video, and that the error resilience techniques provided within the MPEG 2 standard are not sufficient to provide acceptable quality with acceptable overheads, but that this quality can be significantly increased by the addition of a small number of simple techniques. I. INTRODUCTION I N recent years, there has been a trend toward providing mobility in telecommunications systems. The most obvious example of this has been the rapid growth of cellular telephone networks. One key disadvantage of these networks is that they provide only very limited capacity for each user, often less than 16 kb/s. This limitation will be addressed by a number of wireless local-area networks (WLAN’s) that are currently under development. However, there remains the significant issue of errors introduced by the wireless network, which occur very much more frequently than in fixed networks. These transmission errors can result in both packets of data being completely lost, as well as bit errors being introduced into transmitted data. It is this combination of bit errors and cell losses that makes the problems of providing error-resilient compression coding very different compared to conventional fixed networks. In this paper, we examine the impact of the properties of a 50 Mb/s asynchronous transfer mode (ATM)-based WLAN on Manuscript received February 1, 1996; revised June 1996. This paper was supported by the CSIRO Division of Radiophysics. J. Zhang, M. R. Frater, and J. F. Arnold are with the School of Electrical Engineering, University College, The University of New South Wales, Aus- tralian Defence Force Academy, Canberra, ACT 2600, Australia. T. M. Percival is with the CSIRO Division of Radiophysics, Epping NSW 2121, Australia. Publisher Item Identifier S 0733-8716(97)00059-0. moving picture experts group phase 2 (MPEG 2) compressed video traffic. A brief description of the interface between ATM and video codecs is presented. A number methods for providing error resilience in compressed digital video bitstreams transmitted over cell-based telecommunication net- works, such as ATM, are described. We show results on the effect of transmission errors, which can cause cell loss with probabilities as high as 1%, on the quality of the decoded video at the receiver, and demonstrate how error resilience techniques can be used to improve the quality of service. Most important among the findings is that the error resilience techniques provided within the MPEG 2 standard are not sufficient to provide acceptable quality, but that this quality can be significantly increased by the inclusion of a small number of simple additional techniques. In principle, retransmission of errored cells in the WLAN can be used to overcome all transmission errors, except in a small number of cases where there is no significant signal at the receiver. A service that has a high level of resilience against the effect of transmission errors, however, will still be able to operate in those regions of a building where significant propagation errors occur, without a large amount of additional support from the network, thus reducing the complexity of the WLAN, and so making the network implementation less expensive. It is also likely that simpler algorithms for utilizing antenna diversity could be implemented where services are tolerant of transmission errors. Additionally, for real time conversational services, in which there are tight constraints on the permissible end-to-end delay, retransmission of lost or errored data is often not feasible. The novel aspects of this paper are as follows. 1) Video performance in the presence of cell loss is as- sessed for loss streams based on network error/loss measurements. 2) In previous work, only the MPEG 2 video layer has been considered. In a real audio-visual application, the system layer is used to multiplex different streams, such as video and audio. In our work, the MPEG 2 systems layer has been implemented, and is shown to have a significant effect on the quality of the decoded video. 3) The cell loss probabilities that are examined in this paper range between 10 and 10 . The upper end of this range is very much higher than is normally considered in video error resilience work, but is potentially very important in wireless applications. The paper is organized as follows. In Section II, the ATM WLAN is described, followed by brief descriptions of relevant aspects of the MPEG 2 systems and video standards in 0733–8716/97$10.00  1997 IEEE