Optimal bit allocation for streaming stored MPEG-4 FGS video over wireless channel

Wireless video streaming poses great challenges due to the time-varying characteristics of wireless channels and limited network resources. In this paper, we investigate an optimal bit allocation scheme for streaming the stored MPEG-4 FGS (Fine granular scalability) bit-stream from the video server to the clients over wireless channel, which can dynamically adapt to the time-varying channel conditions. MPEG-4 FGS and RS (Reed-solomon) code have been employed as the source coding and channel coding respectively. For MPEG-4 FGS, the reconstructed image quality is proportional to the number of decodable bits from the enhancement layer. Therefore, to achieve the best reconstructed quality at the decoder under a certain channel condition, we propose a decodable rate maximized criterion to optimally determine the truncate rate of the MPEG-4 FGS bit-stream and the channel coding rate. The base layer and different parts of the enhancement layer of MPEG-4 FGS compressed bit-stream are protected unequally according to their importance degree. Experimental results show that, compared with fixed Unequal error protection (UEP) and fixed Equal error protection (EEP), the proposed method can achieve a better performance under various channel conditions.