Video Transmission for Third Generation Wireless Communication Systems

Journal of Research of the National Institute of Standards and Technology, March-April, 2001 by H. Gharavi, S. M. Alamouti

This paper presents a twin-class unequal protected video transmission system over wireless channels. Video partitioning based on a separation of the Variable Length Coded (VLC) Discrete Cosine Transform (DCT) coefficients within each block is considered for constant bitrate transmission (CBR). In the splitting process the fraction of bits assigned to each of the two partitions is adjusted according to the requirements of the unequal error protection scheme employed. Subsequently, partitioning is applied to the LTU-T H.263 coding standard. As a transport vehicle, we have considered one of the leading third generation cellular radio standards known as WCDMA. A dual-priority transmission system is then invoked on the WCDMA system where the video data, after being broken into two streams, is unequally protected. We use a very simple error correction coding scheme for illustration and then propose more sophisticated forms of unequal protection of the digitized video signals. We show that this strategy results in a significantly higher quality of the reconstructed video data when it is transmitted over time-varying multipath fading channels.

Key words: multimedia communications; third generation mobile systems; WCDMA; wireless video.

Accepted: November 1, 2000

Available online: http://www.nist.gov/jres

1. Introduction

The wireless revolution in the 1 980s was primarily driven by market demands for mobile radio voice communications. The first generation analog wireless communication systems used old radio technologies combined with novel cellular network planning [1] to provide transparent ubiquitous mobile access to users. Initially, the cellular market was a niche market targeted at business people. As the price of the user terminals and services dropped due to an unprecedented exponential increase in demand and related economies of scale, cellular radios became the communications tool of choice for the masses. Today, there are in the neighborhood of 300 million mobile phone users in the world. It is expected that in the next decade the number of users will reach the one billion mark.

The second generation wireless systems were introduced in the 1990s and were primarily an evolutionary step towards improving the capacity of cellular systems through digitization of voice and efficient digital modulation schemes. These systems also provided additional features such as security, short messaging, and circuit-switched data.

We are now in the midst of another evolution planned for deployment in the 2000s. This evolution is once again driven by the need for greater bandwidth in anticipation of further demand for voice services. However, the new wireless standards will also provide a pipeline for broadband services such as enhanced high rate data and multimedia services. The global growth of interest in the Internet and in digitized audio and video, and the demand for such services through the fixed communications networks, is an important factor. Although there are currently no applications with great mass market appeal that require broadband wireless access, it is anticipated that the popularity of these services in fixed networks will eventually impact the market for wireless communications.

The next generation wireless systems are required to have voice services of wireline quality and to provide high bit rate data services of 144 kbit/s to 2 Mbit/s depending on the radio environment. At the same time, they are to operate reliably in different types of environments: macro, micro, and pico cellular; urban, suburban, and rural; indoor and outdoor. In other words, the next generation systems are supposed to have better quality and coverage, be more power and bandwidth efficient, and be deployed in diverse environments. These high data rates make video transmission possible for a number of applications such as video conferencing, emergency medical services, and site surveys. However, since most existing video compression standards [7,8,9,10,11] have been developed for relatively benign, nearly error-free environments, they cannot be directly applied in a hostile mobile domain. This is mainly due to the extensive employment of variable length coding techniques, which are efficient in bitrate reductio n terms, but are error-sensitive. A single transmission error may result in an undecodable string of bits. One effective method of protecting the compressed video signal is to split the coded video signal into a number of separate bitstreams where each can be transmitted via a separate channel having a different degree of error protection [12,13,14]. The bitstream splitting can be accomplished by taking into consideration the perceptual significance of coded video, where better protection is provided for the transmission of the perceptually more important bits. In this paper, such a strategy has been considered for one of the leading third generation cellular radio standards known as WCDMA.

The paper begins with an overview of the WCDMA radio standard [2]. Then, a video partitioning scheme based on the ITU-T H.263 standard, is presented. This is followed by a very simple error protection coding scheme for the transmission of partitioned video over IMT-2000 channels. The transmission system model and its parameters are then discussed. This is followed by simulation results that evaluate the transmission of partitioned video using the downlink WCDMA physical layer as a transport vehicle. Finally, for possible future investigations, more sophisticated forms of unequal protection for the digitized video signals are discussed.