Generalized Gaussian Multiterminal Source Coding: The Symmetric Case

Consider a generalized multiterminal source coding system, where <inline-formula> <tex-math notation="LaTeX">$\binom{\ell }{ m}$ </tex-math></inline-formula> encoders, each observing a distinct size-<inline-formula> <tex-math notation="LaTeX">$m$ </tex-math></inline-formula> subset of <inline-formula> <tex-math notation="LaTeX">$\ell $ </tex-math></inline-formula> (<inline-formula> <tex-math notation="LaTeX">$\ell \geq 2$ </tex-math></inline-formula>) zero-mean unit-variance exchangeable Gaussian sources with correlation coefficient <inline-formula> <tex-math notation="LaTeX">$\rho $ </tex-math></inline-formula>, compress their observations in such a way that a joint decoder can reconstruct the sources within a prescribed mean squared error distortion based on the compressed data. The optimal rate-distortion performance of this system was previously known only for the two extreme cases <inline-formula> <tex-math notation="LaTeX">$m=\ell $ </tex-math></inline-formula> (the centralized case) and <inline-formula> <tex-math notation="LaTeX">$m=1$ </tex-math></inline-formula> (the distributed case), and except when <inline-formula> <tex-math notation="LaTeX">$\rho =0$ </tex-math></inline-formula>, the centralized system can achieve strictly lower compression rates than the distributed system under all non-trivial distortion constraints. Somewhat surprisingly, it is established in the present paper that the optimal rate-distortion performance of the afore-described generalized multiterminal source coding system with <inline-formula> <tex-math notation="LaTeX">$m\geq 2$ </tex-math></inline-formula> coincides with that of the centralized system for all distortions when <inline-formula> <tex-math notation="LaTeX">$\rho \leq 0$ </tex-math></inline-formula> and for distortions below an explicit positive threshold (depending on <inline-formula> <tex-math notation="LaTeX">$m$ </tex-math></inline-formula>) when <inline-formula> <tex-math notation="LaTeX">$\rho > 0$ </tex-math></inline-formula>. Moreover, when <inline-formula> <tex-math notation="LaTeX">$\rho > 0$ </tex-math></inline-formula>, the minimum achievable rate of generalized multiterminal source coding subject to an arbitrary positive distortion constraint <inline-formula> <tex-math notation="LaTeX">$d$ </tex-math></inline-formula> is shown to be within a finite gap (depending on <inline-formula> <tex-math notation="LaTeX">$m$ </tex-math></inline-formula> and <inline-formula> <tex-math notation="LaTeX">$d$ </tex-math></inline-formula>) from its centralized counterpart in the large <inline-formula> <tex-math notation="LaTeX">$\ell $ </tex-math></inline-formula> limit except for possibly the critical distortion <inline-formula> <tex-math notation="LaTeX">$d=1-\rho $ </tex-math></inline-formula>.

[1]  Aaron D. Wyner,et al.  The rate-distortion function for source coding with side information at the decoder , 1976, IEEE Trans. Inf. Theory.

[2]  Sui Tung,et al.  Multiterminal source coding (Ph.D. Thesis abstr.) , 1978, IEEE Trans. Inf. Theory.

[3]  Tsachy Weissman,et al.  Multiterminal Source Coding Under Logarithmic Loss , 2011, IEEE Transactions on Information Theory.

[4]  Yasutada Oohama,et al.  Indirect and Direct Gaussian Distributed Source Coding Problems , 2014, IEEE Transactions on Information Theory.

[5]  Jun Chen,et al.  Generalized Gaussian multiterminal source coding and probabilistic graphical models , 2017, 2017 IEEE International Symposium on Information Theory (ISIT).

[6]  Jun Chen,et al.  Vector Gaussian Two-Terminal Source Coding , 2013, IEEE Transactions on Information Theory.

[7]  R. A. McDonald,et al.  Noiseless Coding of Correlated Information Sources , 1973 .

[8]  Toby Berger,et al.  Successive Coding in Multiuser Information Theory , 2007, IEEE Transactions on Information Theory.

[9]  Y. Oohama Gaussian multiterminal source coding , 1995, Proceedings of 1995 IEEE International Symposium on Information Theory.

[10]  Pramod Viswanath,et al.  Rate Region of the Quadratic Gaussian Two-Encoder Source-Coding Problem , 2006, ISIT.

[11]  Thomas M. Cover,et al.  Elements of Information Theory , 2005 .

[12]  Jun Chen,et al.  Vector Gaussian Multiterminal Source Coding , 2012, IEEE Transactions on Information Theory.

[13]  Robert M. Gray,et al.  Toeplitz and Circulant Matrices: A Review , 2005, Found. Trends Commun. Inf. Theory.

[14]  Jun Chen,et al.  On the Sum Rate of Gaussian Multiterminal Source Coding: New Proofs and Results , 2010, IEEE Transactions on Information Theory.

[15]  Zixiang Xiong,et al.  A New Sufficient Condition for Sum-Rate Tightness in Quadratic Gaussian Multiterminal Source Coding , 2013, IEEE Transactions on Information Theory.