Chapter 1
Introduction
1.1 Historical Overview
Multimedia transmission over time-varying wireless channels presents a number of challenges beyond the existing capabilities of Third-Generation (3G) networks. The prevalent design principles stemming from Shannon’s source and channel separation theorem [1] have to be reconsidered. The separation theorem, stating that optimal performance bounds of source and channel decoding can be approached as closely as desired by independently designing source and channel coding, holds only under asymptotic conditions, where both codes may have a length and a complexity tending to infinity, and under the conditions of stationary sources. In practice, the design of the system is heavily constrained in terms of both complexity and delay, and hence source and channel codecs are typically suboptimal [2]. Fortunately, the ‘lossy’ nature of the audio and video codecs may be concealed with the aid of perceptual masking by exploiting the psycho-acoustic and psycho-visual properties of the human ear and eye. The output of practical source encoders, such as speech codecs [3] and image/video codecs [4], may contain a significant amount of redundancy. Moreover, having a nonzero residual channel error rate may lead to dramatically increased source symbol error rates [2]. In these circumstances, the assumptions of Shannon’s separation theorem no longer hold; neither can they be used as a good approximation. One may potentially improve the attainable performance ...
