Bit allocation algorithm for joint source-channel coding of t + 2D video sequences

A joint source-channel coding scheme of t + 2D decomposed video sequences and an iterative bit allocation are presented. The joint source-channel coding scheme consists of a vector quantization and a linear labelling by lattice constellations minimizing simultaneously the channel and the source distortion. The channel distortion, due to the linear labelling, is minimized and depends on the value of the noise variance and the variance of the source. The iterative algorithm results in an optimal codebook allocation subject to a global bit rate and a nonnegativity constraint. The overall flexible coding scheme is proved to be very efficient in noisy environments.

[1]  W. W. Peterson,et al.  Bit Allocation and Encoding for Vector Sources , 1976 .

[2]  Jean-Claude Belfiore,et al.  Algebraic tools to build modulation schemes for fading channels , 1997, IEEE Trans. Inf. Theory.

[3]  X. Giraud,et al.  Optimal linear labelling for the minimization of both source and channel distortion , 2000, 2000 IEEE International Symposium on Information Theory (Cat. No.00CH37060).

[4]  Erik Agrell,et al.  The Hadamard transform-a tool for index assignment , 1996, IEEE Trans. Inf. Theory.

[5]  G. G. Stokes.  "J." , 1890, The New Yale Book of Quotations.

[6]  Yair Shoham,et al.  Efficient bit allocation for an arbitrary set of quantizers [speech coding] , 1988, IEEE Trans. Acoust. Speech Signal Process..